LATE WOODLAND SETTLEMENT AND
SUBSISTENCE IN THE EASTERN UPPER
PENINSULA OF MICHIGAN
By
Sean Barron Dunham

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Anthropology - Doctor of Philosophy
2014

ABSTRACT
LATE WOODLAND SETTLEMENT AND
SUBSISTENCE IN THE EASTERN UPPER
PENINSULA OF MICHIGAN
By
Sean Barron Dunham
This research revisits the debate surrounding Late Woodland subsistence practices in
Michigan’s Upper Peninsula. The Late Woodland period in the Upper Great Lakes region (ca.
A.D. 600 to 1600) is often characterized through models emphasizing the intensive use of a
single, primary key resource, particularly maize, fall spawning fish, or wild rice. For example,
current Late Woodland subsistence models for northern Michigan focus on the intensive harvest,
creation of surplus, and consequent storage of fall spawning fish as the cornerstone of the
settlement and subsistence strategy. New data suggests that the dominant settlement and
subsistence model is incomplete, lacks explanatory value, and requires revision. This study tests
the hypothesis that a suite of potential resources was both present and utilized, allowing for a
more flexible set of strategies, i.e. one based upon multiple rather than a single primary resource.
Archaeological evidence, ethnographic data, and pilot study results reveal that acorns, maize, and
wild rice are likely resources to be incorporated into such a strategy; all can be harvested and
stored in the late summer or fall as a buffer against a poor fish harvest. Each, however, also has
spatial, environmental, and temporal constraints with implications bearing on archaeological site
locations as well as the evidence from the sites themselves.
A spatial analysis of site locations and resource distributions, as well as the composition
of site assemblages was conducted to determine what relationships, if any, can be found between
resources and site locations. The results identified site location patterns relating to the

exploitation of fish as well the potential use of wild rice and acorns, and also revealed changing
patterns of site location over time including an emphasis on coastal settings in the early Late
Woodland and an increase in interior setting sites in the late Late Woodland. In addition, the
study examines strategies for subsistence risk buffering and decision making by Late Woodland
peoples and provides new perspectives on resource scheduling, patterns of mobility, social
organization, and social interaction.
The nature of the data sets employed in the research, as well as the temporal and spatial
scales involved led to the adoption of Resilience Theory as an organizing framework for this
study. The application of Resilience Theory is relatively new in archaeology and in this case
provides a useful contribution to this line of scholarship in a context which has need of greater
theoretical diversity. While an important outcome of the research is a synthesis of our current
understanding of the regional Late Woodland, it also contributes a robust understanding of the
interaction of hunter-gatherers/marginal horticulturalists with their environment.

Copyright by
SEAN BARRON DUNHAM
2014

ACKNOWLEDGEMENTS
This dissertation could not have been completed without the support of numerous
individuals and institutions. While it is impossible to single out everyone who provided some
sort of support, there are some who deserve specific mention. I could not have completed this
work without my guidance committee and I would like to thank Jon Burley, Robert Hitchcock,
Jodie O’Gorman, Ethan Watrall, and especially William Lovis, for their support, advice, and
prodding throughout the process. Much of the data used in this project came from cultural
resource surveys conducted on the Hiawatha National Forest (HNF) and many of these were
carried out through contracts with Commonwealth Cultural Resource Group, Inc. (CCRG). I
owe debts of gratitude to many at CCRG including Donald Weir for allowing me to direct many
of these contract projects, to James Montney for preparing many of the graphics that appear in
this volume, and Michael Hambacher for our numerous discussions on the nuances of UP
archaeology. I also need to express my deep appreciation to John Franzen and Eric Drake of the
HNF for indulging my research as well as for sharing their wealth of knowledge concerning the
archaeology of the north woods. Finally, I am indebted to my family for all their support
through this arduous process: Thanks Mom, Dad, Caroline, Elaine, Paige, and Barron, and
especially Jenn – all my love.

v

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................................ix
LIST OF FIGURES ....................................................................................................................xiv
1.0 INTRODUCTION ..................................................................................................................1
1.1 Research Questions and Context ...................................................................................2
1.1.1 Research Questions .........................................................................................2
1.1.2 Research Context ............................................................................................4
1.2 Theoretical Orientation ..................................................................................................7
1.3 Dissertation Organization ..............................................................................................9
2.0 ENVIRONMENTAL AND CULTURAL BACKGROUND .................................................13
2.1 Environmental Background ..........................................................................................13
2.2 Cultural Background ....................................................................................................17
2.2.1 General Overview of the Woodland Period in the Eastern UP ......................17
2.2.2 The Late Woodland Period............................................................................22
2.3 Background Discussion ................................................................................................31
3.0 ARCHAEOLOGICAL ASSEMBLAGES ..............................................................................34
3.1 Chipped and Ground Stone Artifacts ...........................................................................37
3.1.1 Chipped Stone Tools ....................................................................................37
3.1.2 Chipped Stone Debitage ..............................................................................39
3.1.3 Ground Stone Tools ......................................................................................41
3.2 Late Woodland Ceramics .............................................................................................41
3.3 Other Artifact Classes ..................................................................................................45
3.4 Diversity Use Index .....................................................................................................47
3.5 Subsistence Remains ....................................................................................................54
3.5.1 Faunal Assemblages .....................................................................................56
3.5.2 Floral Remains .............................................................................................74
3.6 Discussion ...................................................................................................................78
4.0 LATE WOODLAND ARCHAEOLOGICAL SITE PREDICTIVE MODEL .......................82
4.1 Development of the Model ..........................................................................................83
4.1.1 Distance to Water ..........................................................................................90
4.1.2 Elevation ......................................................................................................91
4.1.3 Slope ............................................................................................................91
4.1.4 Aspect ..........................................................................................................91
4.1.5 Growing Days ..............................................................................................92
4.1.6 Distance to Potential Wild Rice Stands ........................................................93
4.1.7 Pre-1800 Vegetation .....................................................................................95
vi

4.1.8 Habitat Classification System ......................................................................96
4.1.9 Site Setting ...................................................................................................97
4.1.10 Random Points ............................................................................................98
4.2 Statistical Analysis .......................................................................................................99
4.2.1 t-Test .............................................................................................................99
4.2.2 Kolmogorov-Smirnov Test .........................................................................100
4.2.3 Chi-Square test ...........................................................................................102
4.3 Quadrat Analysis ........................................................................................................105
4.4 Review of the Variables .............................................................................................112
4.4.1 Site Setting ................................................................................................112
4.4.2 Distance to Water ......................................................................................113
4.4.3 Elevation ....................................................................................................114
4.4.4 Slope .........................................................................................................114
4.4.5 Habitat .......................................................................................................115
4.4.6 Wild Rice ..................................................................................................116
4.5 Model Construction ...................................................................................................117
4.6 Review of the Model ..................................................................................................118
4.6.1 Application of the Model ...........................................................................123
4.6.2 East Unit Test .............................................................................................129
4.6.3 Discussion ..................................................................................................132
4.7 Application of the Model to the HNF .......................................................................134
4.8 Discussion .................................................................................................................138
5.0 DIVERSITY USE INDEX AND LATE WOODLAND SITE PREDICTIVE MODEL .....141
5.1 Combining the Diversity Use Index and Late Woodland Site Predictive Model ......141
5.2 Late Woodland Site Predictive Model and Habitat Diversity Catchments................150
5.2.1 150 m Radius Catchments ..........................................................................151
5.2.2 150 m Radius Catchments and DUI............................................................158
5.2.3 1500 m Radius Catchments ........................................................................160
5.2.4 1500 m Radius Catchments and DUI..........................................................168
5.3 Discussion ................................................................................................................170
6.0 CONCLUSIONS.................................................................................................................176
6.1 Fish, Acorns, Wild Rice, and Maize .........................................................................177
6.1.1 Fall Fishery .................................................................................................178
6.1.2 Maize...........................................................................................................181
6.1.3 Wild Rice ...................................................................................................184
6.1.4 Acorns .........................................................................................................187
6.1.5 Discussion ...................................................................................................189
6.2 Extended Diversity Sites, Persistent Places, and Anthropomorphic Landscapes ......193
6.3 Extended Diversity Sites and Ceramics ....................................................................199
6.4 Concluding Thoughts .................................................................................................205
APPENDICES .............................................................................................................................209
Appendix A: Eastern UP Landscape Ecosystems ...........................................................210
Appendix B: Baseline Site Data ....................................................................................215
vii

Appendix C: Additional Site Data ..................................................................................221
Appendix D: Chipped and Ground Stone Artifacts ........................................................224
Appendix E: Ceramics ....................................................................................................231
Appendix F: Other Artifacts ...........................................................................................238
Appendix G: DUI Rank ..................................................................................................240
Appendix H: Fauna .........................................................................................................243
Appendix I: Flora ............................................................................................................248
Appendix J: HNF Site Data ............................................................................................252
Appendix K: Archaeological Survey on the HNF ..........................................................255
Appendix L: Variables and Summary Statistics .............................................................262
Appendix M: Pre-1800 Vegetation Coding ....................................................................267
Appendix N: Habitat Classification Coding ...................................................................270
Appendix O: Regression Analysis Output ......................................................................296
Appendix P: Site Predictive Scores ................................................................................302
Appendix Q: Site Data (LWPM and DUI Ranks) ..........................................................306
Appendix R: 150 m Radius Catchment Data (SS&HD and DUI Ranks) .......................309
Appendix S: 1500 m Radius Catchment Data (SS&HD and DUI Ranks) .....................311
Appendix T: Fall and Spring Spawning Fish at Extended and Intermediate Diversity
Sites ...........................................................................................................313
Appendix U: Sites Associated with Wild Rice Locales..................................................315
Appendix V: Sites in Mixed Pine Habitats .....................................................................317
Appendix W: Extended Diversity Sites ..........................................................................319
Appendix X: Ceramics Data ...........................................................................................321
BIBLIOGRAPHY ...................................................................................................................... 323

viii

LIST OF TABLES

Table 1: Mean number of identified fauna and mean diversity score...........................................59
Table 2: Fall Spawning Fish Recovered from Coastal Sites.........................................................67
Table 3: Habitat and seasonal summary for the most ubiquitous fish. .........................................72
Table 4: Habitat and seasonal summary for the most ubiquitous mammals.................................73
Table 5: Pearson’s Correlation Coefficient...................................................................................99
Table 6: Summary of t-Tests.......................................................................................................100
Table 7: Summary of Kolmogorov-Smirnoff (KS) Test of Site Frequency Distribution. .........101
Table 8: Summary of Kolmogorov-Smirnoff (KS) Test of Site Frequency Distribution. .........102
Table 9: Summary of Kolmogorov-Smirnoff (KS) Tests. ..........................................................102
Table 10: Summary of Chi-Square Tests. ...................................................................................103
Table 11: Elevation in Feet amsl. ...............................................................................................108
Table 12: Slope in Degrees. ........................................................................................................108
Table 13: Aspect in Degrees. ......................................................................................................109
Table 14: Growing Days. ............................................................................................................109
Table 15: Mean Distance to Water and Habitat Diversity. .........................................................110
Table 16: Wild Rice in West Unit Quadrats (Yates Correction). ..............................................111
Table 17: Late Woodland Site Predictive Model Point Scales. .................................................119
Table 18: Late Woodland Site Archaeological Predictive Scores. ............................................119
Table 19: West Unit Late Woodland Archaeological Sensitivity. .............................................120
Table 20: Eastern UP Late Woodland Archaeological Sensitivity. ...........................................120
Table 21: Summary Statistics Comparing Archaeological Predictive Scores. ..........................121

ix

Table 22: Summary of Chi-Square Tests. ...................................................................................121
Table 23: Summary Statistics by Site Setting. ............................................................................122
Table 24: Summary Statistics Comparing Test Areas. ...............................................................125
Table 25: Comparison of Late Woodland Archaeological Sensitivity. .....................................126
Table 26: Comparison of Late Woodland Archaeological Sensitivity. .....................................126
Table 27: Archaeological Sensitivity by HNF Unit and Overall. ..............................................137
Table 28: Likelihood of Encountering a Late Woodland Site. ...................................................137
Table 29: Late Woodland sites per Diversity Use Index (DUI) and Site Predictive Model
(LWPM). ....................................................................................................................142
Table 30: Early and Late Late Woodland (LW) sites per Diversity Use Index (DUI) and Site
Predictive Model (LWPM). .......................................................................................147
Table 31: Early and Late Late Woodland (LW) sites per Diversity Use Index (DUI) and Site
Predictive Model (LWPM). .......................................................................................148
Table 32: Descriptive Statistics of the 150 m Radius Catchments for Random Points and Late
Woodland Sites Addressing Mean Predictive Score (MPS) and Mean Habitat
Diversity Index (MDHI). ...........................................................................................152
Table 33: Results to t-Tests for 150 m Radius Catchments. .......................................................152
Table 34: Descriptive Statistics for 150 m Radius Catchments Addressing Mean Predictive
Score (MPS), Mean Habitat Diversity Index (MDHI), and Catchment Area. ...........154
Table 35: Descriptive Statistics for 150 m Catchments Addressing Mean Predictive Score
(MPS), Mean Habitat Diversity Index (MDHI) for Interior Sites. ...........................157
Table 36: Results to t-Tests for 1500 m Radius Catchments. .....................................................162
Table 37: Descriptive Statistics of the 1500 m Radius Catchments for Random Points and Late
Woodland Sites Addressing Mean Predictive Score (MPS) and Mean Habitat
Diversity Index (MDHI). ...........................................................................................162
Table 38: Descriptive Statistics for 1500 m Radius Catchments Addressing Mean Predictive
Score (MPS), Mean Habitat Diversity Index (MDHI), and Catchment Area. ...........165
Table 39: Descriptive Statistics for 1500 m Catchments Addressing Mean Predictive Score
(MPS), Mean Habitat Diversity Index (MDHI) for Interior Sites. ............................168

x

Table 40: Estimation of the number of red oak trees per hectare in Mixed Pine Habitats (after
Price 1994). ................................................................................................................190
Table 41: Baseline Site Data .......................................................................................................216
Table 42: Features .......................................................................................................................219
Table 43: Chronometric Ages .....................................................................................................220
Table 44: Chipped Stone Tools...................................................................................................225
Table 45: Debitage ......................................................................................................................227
Table 46: Ground Stone Tools ....................................................................................................230
Table 47: Baseline Ceramic Data ...............................................................................................232
Table 48: Late Woodland Ceramic Types (MNV) .....................................................................234
Table 49: Oneota-Related and Other Ceramic Types (MNV) ....................................................236
Table 50: Other Artifact Types ...................................................................................................239
Table 51: Diversity Use Index (DUI) Rank ................................................................................241
Table 52: Fauna from sites within the HNF (Identified Species Only) ......................................244
Table 53: Fauna from Sites outside the HNF (Identified Species Only) ....................................245
Table 54: Faunal Diversity Index ...............................................................................................247
Table 55: Botanical Remains (PEB Identified/Carbonized Only) ..............................................249
Table 56: Wood Charcoal Remains (PEB Identified/Carbonized Only) ....................................251
Table 57: HNF Site Data.............................................................................................................253
Table 58: Variables and Summary Statistics ..............................................................................263
Table 59: Variables and Summary Statistics LW Sites ..............................................................264
Table 60: Variables and Summary Statistics Random Points.....................................................265
Table 61: Pre-1800 Vegetation Coding ......................................................................................268
Table 62: Alger County ..............................................................................................................271
xi

Table 63: Chippewa County .......................................................................................................277
Table 64: Delta County ...............................................................................................................281
Table 65: Luce County................................................................................................................284
Table 66: Mackinac County ........................................................................................................288
Table 67: Schoolcraft County .....................................................................................................292
Table 68: Categorical Variables..................................................................................................297
Table 69: Calculation Parameters ...............................................................................................297
Table 70: Sample Split ................................................................................................................297
Table 71: Log-Likelihood Iteration History ...............................................................................298
Table 72: Information Criteria ....................................................................................................298
Table 73: Parameter Estimates....................................................................................................299
Table 74: Odds Ratio Estimates ..................................................................................................299
Table 75: Overall Model Fit .......................................................................................................299
Table 76: R-Square Measures .....................................................................................................299
Table 77: Model Prediction Success Table .................................................................................300
Table 78: Summary of Prediction Success Table .......................................................................300
Table 79: Simulation Vector .......................................................................................................301
Table 80: Predictive Scores HNF Sites .......................................................................................303
Table 81: Predictive Scores Other Sites......................................................................................305
Table 82: Site Data (LWPM and DUI Ranks) ............................................................................307
Table 83: 150 m Radius Catchment Data (SS&HD and DUI Ranks) ........................................310
Table 84: 1500 m Radius Catchment Data (SS&HD and DUI Ranks) ......................................312
Table 85: Extended Diversity Sites.............................................................................................314
xii

Table 86: Intermediate Diversity Sites .......................................................................................314
Table 87: Sites Associated with Wild Rice Locales ...................................................................316
Table 88: Sites in Mixed Pine Habitats.......................................................................................318
Table 89: Extended Diversity Sites.............................................................................................320
Table 90: Ceramics Data ...........................................................................................................321

xiii

LIST OF FIGURES

Figure 1: Location of the Eastern Upper Peninsula of Michigan. .................................................2
Figure 2: Timeline showing selected cultural and physical environmental variables relating to
the EUP. .........................................................................................................................................15
Figure 3: River Drainages in the Eastern Upper Peninsula. ........................................................17
Figure 4: Locations of the 81 Late Woodland Archaeological Sites. ..........................................35
Figure 5: Percentage of ceramics by type (North/South). ............................................................43
Figure 6: Percentage of ceramics by type (East/West). ...............................................................44
Figure 7: Percentage of ceramics by age (North/South). .............................................................45
Figure 8: Percentage of ceramics by age (East/West). .................................................................45
Figure 9: Scatter Plot DUI and DUIrev scores. ............................................................................51
Figure 10: The Locations of Late Woodland Sites with Extended Diversity. .............................52
Figure 11: LW components with identifiable fauna. ....................................................................57
Figure 12: Faunal Diversity Score. ..............................................................................................58
Figure 13: Best represented and other notable species. ...............................................................61
Figure 14: Relative percentage of each of the best represented species by coastal & interior
setting. .........................................................................................................................63
Figure 15: Relative percentage of the most ubiquitous species comparing early LW and late
LW components. .........................................................................................................64
Figure 16: Relative percentage of the most ubiquitous species by east and west. .......................65
Figure 17: Number of sites with fall spawning fish by broad region. .........................................68
Figure 18: Number of sites with spring and fall spawning fish. ..................................................69
Figure 19: Comparison of fall spawning fish, spring spawning fish, & big game by count. ......70
Figure 20: Comparison of fall spawning fish, spring spawning fish, & big game by age. ..........70
xiv

Figure 21: Diversity of floral assemblages (seed/nut and charcoal) by site. ...............................75
Figure 22: Percentage of sites by category with floral remains, faunal remains, or subsurface
features. .....................................................................................................................80
Figure 23: Hiawatha National Forest. ...........................................................................................83
Figure 24: GIS Model Flowchart. .................................................................................................87
Figure 25: Late Woodland Site Locations in the Hiawatha National Forrest. ..............................88
Figure 26: Growing Days..............................................................................................................93
Figure 27: Location of Wild Rice Patches. ...................................................................................95
Figure 28: Random Points on the West Unit of Hiawatha National Forest. .................................98
Figure 29: West Unit Quadrats. ..................................................................................................106
Figure 30: Comparison of Quadrat Habitats. ..............................................................................111
Figure 31: Archaeological Sensitivity of Coastal Late Woodland Site Locals by Great Lake...123
Figure 32: Location of the Randomly Selected West Unit Quadrats..........................................125
Figure 33: Location of the 2011 Survey Areas. ..........................................................................127
Figure 34: Quadrat W10. ............................................................................................................128
Figure 35: 2011 Survey Area 2. ..................................................................................................129
Figure 36: Location of the Randomly Selected East Unit Quadrats. ..........................................130
Figure 37: 2012 Survey Area Locations. ....................................................................................131
Figure 38: Quadrats W6 and W7. ...............................................................................................133
Figure 39: West Unit Archaeological Sensitivity. ......................................................................135
Figure 40: East Unit Archeological Sensitivity. .........................................................................136
Figure 41: Comparison of Diversity Use Index and Late Woodland Site Predictive Model by
Category (percent). ....................................................................................................143

xv

Figure 42: Comparison of Diversity Use Index and Late Woodland Site Predictive Model by
Category and Site Setting (percent). .........................................................................144
Figure 43: Comparison of Diversity Use Index and Late Woodland Site Predictive Model by
Category and Interior Site Setting (count). .............................................................145
Figure 44: Comparison of Diversity Use Index and Late Woodland Site Predictive Model by
Category and Relative Age (percent). .....................................................................147
Figure 45: Comparison of Diversity Use Index and Late Woodland Site Predictive Model
Coastal Sites by Category and Relative Age (percent). ..........................................149
Figure 46: Comparison of the habitat composition between Random Points and Late
Woodland sites in the 150 m radius catchments (percent). .....................................153
Figure 47: Map showing the locations of the 150 m radius Late Woodland site clusters as well
as selected Late Woodland sites noted in text. ........................................................154
Figure 48: Scatter plot comparing coastal and interior LW sites by Mean Predictive Score
(MPS) and Mean Habitat Diversity Index (MHDI) for 150 m radius catchments. .155
Figure 49: Comparison of the habitat composition between Coastal and Interior Late
Woodland sites in the 150 m radius catchments (percent). .....................................156
Figure 50: Comparison of the habitat composition between Interior Lake Late Woodland site
catchments and River/Stream Late Woodland 150 m radius site catchments
(percent). ..................................................................................................................158
Figure 51: Comparison of Diversity Use Index and Site Sensitivity & Habitat Diversity by
Category in the 150 m Radius Site Catchments (percentage). ................................160
Figure 52: Comparison of Diversity Use Index and Site Sensitivity & Habitat Diversity by
Category and Site Setting, 150 m Radius Site Catchments (percentage). ...............161
Figure 53: Comparison of Diversity Use Index and Site Sensitivity & Habitat Diversity by
Category and Interior Site Setting, 150 m Radius Site Catchments (count). ..........161
Figure 54: Comparison of the habitat composition between Random Points and Late
Woodland sites in the 1500 m radius catchments (percent). ...................................163
Figure 55: Map showing the locations of the 1500 m radius Late Woodland site clusters as
well as selected Late Woodland sites noted in text. ................................................164
Figure 56: Scatter plot comparing coastal and interior LW sites by Mean Predictive Score
(MPS) and Mean Habitat Diversity Index (MHDI) for 1500 m radius catchments. 166

xvi

Figure 57: Comparison of the habitat composition between Coastal and Interior Late
Woodland site 1500 m radius site catchments (percent). ........................................167
Figure 58: Comparison of Diversity Use Index and Site Sensitivity & Habitat Diversity by
Category in the 1500 m Radius Site Catchments (percentage). ..............................169
Figure 59: Comparison of Diversity Use Index and Site Sensitivity & Habitat Diversity by
Category and Site Setting, 1500 m Radius Site Catchments (percentage). .............170
Figure 60: Comparison of Diversity Use Index and Site Sensitivity & Habitat Diversity by
Category and Interior Site Setting, 1500 m Radius Site Catchments (count). ........170
Figure 61: Comparison of Diversity Use Index by Relative Age of Coastal Late Woodland
Sites. ........................................................................................................................173
Figure 62: Distribution of Late Woodland sites associated with wild rice habitats. .................185
Figure 63: Percentage of LW Sites with Mixed Pine Habitat.. ...................................................187
Figure 64: Comparison of Coastal/Interior Sites by Percentage of Mixed Pine Habitat. ...........188
Figure 65: Distribution of Late Woodland sites in Mixed Pine Habitats by Diversity Index. ...190
Figure 66: Proportion of High, Medium and Low Sensitivity Areas by Percent. .....................195
Figure 67: Location of Grand Island Sites In Relation to Site Sensitivity. ................................197
Figure 68: Distribution of early Late Woodland Ceramics. ......................................................201
Figure 69: Distribution of late Late Woodland Ceramics. .........................................................202
Figure 70: Distribution of late Late Woodland sites by Diversity Index. ..................................204
Figure 71: Receiver Operating Characteristic Curve ..................................................................300

xvii

1.0

INTRODUCTION
This research revisits the contentious topic and longstanding debate surrounding Late

Woodland (AD 600 to AD 16001) settlement patterns and subsistence practices in the Upper
Great Lakes region (cf., Cleland 1982, 1989; Martin 1985, 1989; B. A. Smith 2004). Previous
research in the eastern Upper Peninsula (hereafter UP) of Michigan has emphasized the Great
Lakes fishery, especially the intensive harvest of spring and fall spawning fish. These models
are primarily based on fieldwork conducted in the 1960s and early 1970s that focused on coastal
sites and, therefore, may be skewing the subsistence trends towards aquatic species. More recent
archaeological studies in the eastern UP have identified and tested numerous Late Woodland
sites in both coastal and interior settings. These data as well as the results of pilot studies
suggest that the existing settlement and subsistence model is incomplete and, therefore, not
tenable. For the purpose of this study, the eastern UP is defined as Alger, Chippewa, Delta,
Luce, Mackinac, and Schoolcraft Counties (Figure 1).
It is well understood that the cold winter in the Upper Great Lakes region poses a
subsistence risk and that one mechanism by which to mitigate this risk is to set a portion of a
resource aside for later use (i.e., storage). Existing models of Late Woodland subsistence focus
on the intensive harvest, the creation of surpluses, and consequent storage of fall spawning fish
as the cornerstone of such a risk buffering strategy (Cleland 1982). There is not a sufficient
number and range of risk buffering mechanisms to allow the fall fishery to serve as an exclusive
resource, although it was certainly a primary resource. This study tests the hypothesis that a
suite of potential resources is present that allows for a more flexible set of alternative risk
buffering strategies. Archaeological, ethnographic, and ethnohistoric data as well as the results
1

Dates are presented in Western calendar format (BC/AD) or in years before present (BP) unless otherwise
specified.

1

Figure 1: Location of the Eastern Upper Peninsula of Michigan.

of pilot studies provide evidence that acorns, maize, and wild rice are likely candidates for
alternate resources that could be harvested and stored in the late summer or fall as a buffer
against a poor fish harvest (Dunham 2008; 2009; O’Shea 2003).

1.1

Research Questions and Context

1.1.1 Research Questions
Based on the current models, a number of technological and social changes occur during
the course of the Woodland period in the Upper Great Lakes region (AD 1 – AD 1600) that
culminates in the development of the fall fishery (Cleland 1982; Martin 1985; McHale-Milner
1991). The current models and the processes behind them are more fully explored in Chapter
2

2.2, but it is important to highlight the fall fishery as critical to the research questions behind this
study. The harvest of fall spawning fish became possible with the adoption of the gill net.2 The
gill net also facilitated the collection of high volumes of fall spawning lake trout and white fish
which could be processed and stored in surplus and used to offset the diminished availability of
resources over the cold UP winter.
The overarching question addressed in the current study is whether or not the
development of the fall fishery provided a sufficient buffer to offset winter resource shortfalls. It
is hypothesized that the integration of one or more alternate, highly productive food resources
into the system would serve as a further buffer against such shortfalls as well as a buffer against
poor fish harvests (see O’Shea 1989). The pilot studies mentioned above have shown that
acorns, wild rice, or maize could have filled this niche, singly or in concert, creating an
additional buffer (Dunham 2008, 2009; O’Shea 2003).
The adoption of the fall fishery is also viewed as a catalyst for social change in the Late
Woodland period (Cleland 1982; McHale-Milner 1991; see also Chapter 2.2). The primary
outcomes of this are perceived as increased population, as a result of diminished winter risk and
more consistent food supply, and greater social organization and social integration brought about
through the need to coordinate and schedule the fall fishery as well as to process the harvest of
the high volume of fish (see also Braun and Plog 1982; Brown 1985; Schalk 1977; for an
alternate perspective see S. Martin 1985; 1989).
Since the work of Cleland (1982) and S. Martin (1985) on this topic, there has been a
significant amount of archaeological investigation in the eastern UP, largely driven by federally
mandated compliance projects, that has expanded the body of data available to explore these

2

A gill-net is defined as a “long, coarse, mesh net set to form an underwater ‘curtain’ in which fish become
ensnared by their gills” (Cleland 1982:774).

3

questions. The results of these investigations have already begun to reshape the narrative
concerning the development of the fall fishery in the Upper Great Lakes (see especially B. A.
Smith 2004; see also Chapter 2.2]). Additionally, these projects have begun to reveal a broader
settlement system that includes Late Woodland sites in the interior (Dunham 2002; Franzen
1987; S. Martin 1999).
In summary, the following questions will be asked throughout this study: 1) did Late
Woodland people in the eastern UP use maize, acorns, and/or wild rice as a buffer against a poor
yield of fall spawning fish?; 2) if so, what is the evidence to support the use of these resources?;
3) if not, were there other buffering mechanisms used to offset a poor fishing season? (see also
Chapter 2 for additional context); 4) is there additional archaeological evidence to support the
changing social dynamics proffered in the current models? (see Chapter 2.2 for additional
context; and 5) what was the role of interior resources in the Late Woodland period in the
eastern UP or why does it appear that there was greater use of the interior by Late Woodland
people?
1.1.2 Research Context
The relationship between hunter-gatherers and their physical environment is a critical
facet in the study of hunter-gatherers (Bettinger1987; Jochim 1991; Kelly 1995). An important
theme throughout this dissertation is the interaction between humans and their environment.
According to B. D. Smith (2007:1797), “Many animal species attempt to enhance their
environments, and humans have been trying to make the world a better place to live – for
themselves – for tens of thousands of years, often with unforeseen consequences.” It is through
their niche construction and niche maintenance that humans effectively manage resources, and it
is this relationship which seems to set the stage for many discussions of hunter-gatherer

4

subsistence systems which, in turn, play a significant role in all other aspects of hunter-gatherer
societies (such as social structures, settlement patterns, and demography to name a few).
Access to subsistence resources is critical to hunter-gatherers. However, knowing that a
resource is present and being able to successfully predict its occurrence is as critical to the
discussion as access to that resource. The interaction between hunter-gatherers and their
environment would lead to information about the location of resources as well as information
about the availability of resources (Binford 1983; Davidson-Hunt and Berkes 2003; Kelly 1992;
Whallon 2006). All the previous research concerning the Woodland period in the eastern UP
assumes that the population was mobile and followed a seasonal subsistence round (see Brashler
et al. 2000; Brose and Hambacher 1999; Buckmaster 1979; Franzen 1986; Holman 1978; S.
Martin 1999). The locations of subsistence resources as well as the timing and potential returns
from those resources are critical factors concerning hunter-gatherer mobility (see Binford 1983;
Davidson-Hunt and Berkes 2003; Harpending and Davis 1977; Lovis et al. 2005; Morgan 2009).
Mobility allows greater flexibility for accessing resources, especially if subsistence resources are
aggregated in space and by season as they are in the Late Woodland in the eastern UP (see
Chapter 2.2).
Certain plants and animals follow predictable seasonal patterns that can facilitate their
reliable exploitation by hunter-gatherers. In the prevailing model for the Woodland period in the
Upper Great Lakes region, the seasonal patterns of spring and fall spawning fish illustrates this
well (c.f., Cleland 1982). In each instance, the resource has the potential to produce a high
volume of food over a short period of time which, in turn, facilitates a more intensive
exploitation of that resource. A byproduct of these highly productive systems is a surplus that
aids in creating more reliable resources (see Ingold 1983; O’Shea 1981).

5

This type of pattern has been described as a more intensive use of specific resource
(Binford 2001; Keeley 1995; B. D. Smith 2001). B. D. Smith (2001) explores the concepts of
“intensification” and “resource management.” Each of these terms reflects “deeper and more
complex relationships of interaction with plant and animal communities” (B. D. Smith 2001:35)
Management of resources is a method of facilitating their predictability and reliability
(Hildebrand 2003; Keeley 1995; Miller and Davidson-Hunt 2010). Acorns, wild rice and maize
each represent plant resources that can be, and are, managed by hunter-gatherers and low level
food producers (sensu B. D. Smith 2001).
Storage is also an important factor in resource predictability and reliability. Storage can
be used as a mechanism to stabilize or bridge unpredictable resources and periods of resource
scarcity and shortfall, or as a method to facilitate resource scheduling (Binford 1983; Ingold
1983; O’Shea 1981; see also Bursey 2001; Dunham 2000a). Direct, or practical, storage
involves setting a portion of a resource aside for later use which extends the availability of the
resource and increases its reliability. Social storage, the generation of reciprocal sharing
relationships often offset in time and space, encompasses a range of interactions, but establishes
relationships between individuals and groups (see also Braun and Plog 1982; Holman and Lovis
2008; O’Shea 1981; Parkinson 2002). Resources and information are exchanged through this
network that may provide alternatives in instances where resources are scarce or unreliable (see
also Binford 1983; Davidson-Hunt and Berkes 2003; Whallon 2006).
This study examines strategies for subsistence risk buffering and decision making by Late
Woodland peoples and will provide new perspectives on resource scheduling, patterns of
mobility, social organization, and social interaction. It is posited that the distribution of Late
Woodland archaeological sites in the eastern UP is patterned by the decisions of Late Woodland

6

people in relation to environmental factors they considered culturally and economically
important, specifically the location of subsistence resources. Fall spawning fish, acorns, maize,
and wild rice each have spatial, environmental, and temporal (seasonal and inter-annual)
constraints. Pilot studies as well as previously published research have shown relationships
between each of these resources and their associated constraints (see for example, Cleland 1982;
Dunham 2008, 2009; Franzen 1987; S. Martin 1985; Moffat and Arzigian 2000; O’Shea 2003; B.
A. Smith 2004; Vennum 1988; Yarnell 1964). These constraints have implications concerning
archaeological site locations as well as the archaeological evidence at the sites themselves.

1.2

Theoretical Orientation
The research relies on multiple data sets (cultural and environmental [see also Chapter 2])

that span approximately 1000 years (the Late Woodland period, ca. AD 600 to AD 1600) and
encompass a large geographic area (eastern UP, ca. 6752 square miles [mi²] or 17,488 square
kilometers [km]). The nature of the data as well as the temporal and spatial scales has led to the
adoption of resilience theory as a framework for this study. The underlying element of
Resilience Theory is that nested adaptive cycles can be viewed both synchronically and
diachronically to observe continuity and change across time and/or geographic space (c.f.,
Redman and Kinzig 2003; Walker et al. 2006).
Resilience Theory originated in ecology (Holling 1973; see also Delcourt and Delcourt
2004). Inherent in this perspective is the notion that smaller, faster cycles and larger, slower
cycles may interact and either amplify or dampen different effects to cause change or stability.
Walker et al. (2006:13-2) define resilience as “the capacity of a system to experience shocks
while retaining the same function, structure, feedback, and therefore identity.” Stability within

7

changing cycles can be classified as resilience and change can be viewed as reorganization or
adaptation.
Another important component of Resilience Theory is the role and incorporation of
information exchange and long term memory (Redman and Kinzig 2003; Walker et al. 2006).
This element of Resilience Theory, along with the ability to move across spatial and temporal
scales and multiple social and environmental variables, couples well with archaeology because
of the ability of archaeologists to view the interaction between cultural and ecological systems
over time (Redman and Kinzig 2003). Two recent studies show the utility of Resilience Theory
in understanding the interplay between cultural and physical environments at a regional level and
which explore mobility patterns and subsistence practices in times of environmental
change/instability (Nelson et al. 2006; Thompson and Turck 2009; see also Minc 1986). The
study by Thompson and Turck (2009) is particularly informative in that it addresses cycles of
reorganization and resilience in relation to hunter-gatherer use of coastal environments.
Resilience Theory offers a useful heuristic framework within which to explore
subsistence risk buffering, resource scheduling (including the potential for intensification of
resource use), patterns of mobility, social organization, and social interaction. Such an approach
acknowledges the dynamic nature of cultural processes and is capable of accommodating
seemingly disparate cycles including the likely maintenance (resilience) of subsistence practices,
such as harvesting spring spawning fish or acorns, along with the incorporation (reorganization)
of new resources like fall spawning fish or maize into the system.
Resilience theory also has much in common with elements of social ecological systems,
settlement ecology, and traditional ecological knowledge (Anschuetz et al. 2001; Davidson-Hunt
2000; Davidson-Hunt and Berkes 2003; Jones 2010; Trusler and Johnson 2008). Each of these

8

approaches recognizes that landscapes are products of people’s interaction with the environment.
Information exchange and memory facilitate resilience or adaptation in a cultural system through
knowledge, tradition, and institutions – each of which are critical components of social
ecological systems and traditional ecological knowledge (Davidson-Hunt and Berkes 2003; Funk
2004; Johnson 2000; Miller and Davidson-Hunt 2010; see also Holman and Lovis 2008; Sobel
and Bettles 2000; Whallon 2006). Settlement ecology provides an accessible archaeological
framework addressing “issues of archaeologically observed patterns of land use, occupation, and
transformation over time” (Anschuetz et al. 2001: 177). These perspectives are well suited to
address questions of landscapes and site locations from ecological and cultural factors across
space as well as over time.

1.3

Dissertation Organization
The dissertation research revisits the topic of settlement patterns and subsistence

practices in the Late Woodland period of the eastern UP and evaluated it against the body of data
generated over the past 25 years. If people were using fall spawning fish, acorns, wild rice,
and/or maize, then their sites should be located to access these resources. Previous research in
the Upper Great Lakes region has shown that many Late Woodland coastal sites are well placed
to access fall spawning beds and include archaeological evidence (net sinkers and faunal
remains) which confirm this activity (Cleland 1982; Holman 1978; Lovis and Holman 1976;
Martin 1985; B. A. Smith 2004). In addition to fish habitat, pilot studies have shown potential
relationships between the distribution of Late Woodland archaeological sites and habitats with
potential oak (acorn) and/or wild rice (Dunham 2008; 2009).
A major focus of the dissertation research was to carry out a spatial analysis of Late
Woodland site location as well as an analysis of the composition of site assemblages in an effort
9

to determine what relationships, if any, can be found between resource distribution and site
locations. While an important outcome of the research is a synthesis of our current understanding
of regional Late Woodland archaeological research, it also contributes a more robust
understanding of the interaction of hunter-gatherers with their environment.
The dissertation is organized into six chapters and is supported by references and
multiple appendices. Following this introduction, Chapter 2 provides a review of the
environmental background and cultural setting for the research conducted in this dissertation.
These background sections are meant to provide context for the discussions in subsequent
chapters.
Chapter 3 provides a review and synthesis of the available archaeological assemblage
data for LW sites in the eastern UP. The review relied on previously completed studies and
includes published and unpublished sources including the so-called “gray literature”.3 The
archaeological data in Chapter 3 was summarized in tabular form to establish a format to
compare the characteristics of each assemblage. Much of the archaeological site data that has
been generated over the past 25 years has not been quantified or synthesized beyond the site
specific level. In certain instances, such as the pilot studies cited above, patterns have been
observed with significant implications for our understanding of the prehistory of the region. The
broader regional synthesis of Late Woodland data should provide additional insights that will
enhance our understanding of Late Woodland cultural dynamics.
Eighty-one Late Woodland sites were identified for this study. The data derived from
these sites was collected in a variety of ways, including small scale surface collection, limited
test excavations, and large scale excavations. Further, some of the sites where excavation has

3

In this instance, gray literature refers to the technical reports prepared for federal compliance projects as well as
other unpublished written sources. For a broader discussion of gray literature see Seymour (2010).

10

taken place revealed cumulative palimpsests.4 To address such palimpsests as well as the scale
and types of investigation a diversity use index was created as a measure to compare the various
site assemblages (c.f., Kvamme 1985). This analysis also allowed the sites to be classified into
categories that allowed sites to be characterized by their likely organizational structure (c.f.,
Binford 1980).
An inductive Late Woodland archaeological site predictive model was also created using
the environmental characteristics of the Late Woodland site settings (Chapter 4). This aspect of
the dissertation focused on 48 Late Woodland archaeological sites located in the Hiawatha
National Forest. This approach was taken because of the consistent approach to archaeological
survey used by the Forest as well as the emphasis on conducting archaeological survey in coastal
and interior settings. Additionally, this data set included the highest proportion of interior Late
Woodland sites. The results of this analysis allowed all 81 Late Woodland sites in the eastern
UP to be classified as to the relative archaeological sensitivity of their site location.
Preliminary examinations of the environmental context of the eastern UP in the Late
Woodland period were conducted as part of pilot studies (Dunham 2008; 2009). The pilot
studies have also shown a statistically significant relationship between the location of Late
Woodland archaeological sites in the eastern UP and habitats that were likely to include oak
and/or potentially include wild rice (Dunham 2008; 2009). While these relationships do not
demonstrate that people were using these resources, they offer the potential that people and these
resources were occupying the same geographic space. An important outcome of Chapter 4 was
testing the results of the pilot studies on a finer scale.

4

According to Bailey (2007:204), “a cumulative palimpsest is one in which the successive episodes of deposition, or
layers of activity, remain superimposed one upon the other without loss of evidence, but are so re-worked and mixed
together that it is difficult or impossible to separate them out into their original constituents.”

11

The results of the syntheses of the assemblages, the generation of the diversity use index,
and the creation of the Late Woodland site predictive model were explored to see what trends
might be identified concerning Late Woodland site location and resource use (Chapter 5).
Additionally, an expedient catchment analysis of the site locales was also conducted to examine
the environmental settings of the sites. The resulting analyses revealed trends that directly
inform on the primary questions raised by this study. Late Woodland settlement patterns have
spatial, temporal, and environmental components that can be identified and explored.
Finally, the results of the research and analyses are presented in Chapter 6. This chapter
is divided into four subsections. The first discusses the potential for the use of wild rice, acorns
and maize, as well as the fall fishery, for the Late Woodland period in the eastern UP. It also
explores the temporal and spatial potentials of these resources from the context of site function.
The second subsection discusses the potential for landscape management by Late Woodland
people in the eastern UP and introduces the concept of persistent places to describe a small
subset of the eastern UP Late Woodland sites (c.f., Schlanger 1992; Thompson 2010). The third
subsection continues on the theme of persistent places and explores the distribution of ceramic
types in time and space, and how this may reflect social organization. The final subsection
provides an overview of the patterns observed in the context of continuity and change across
time and space.

12

2.0

ENVIRONMENTAL AND CULTURAL BACKGROUND
The existing Late Woodland subsistence models for northern Michigan focus on the

intensive harvest, creation of surplus, and consequent storage of fall spawning fish as the
cornerstone of the settlement and subsistence strategy. The proposed research revisits the debate
surrounding Late Woodland subsistence practices in Michigan’s Upper Peninsula. New data
suggests that the dominant settlement and subsistence model is incomplete, lacks explanatory
value, and requires revision. This chapter provides a review of the environmental background
and cultural setting for the research conducted in this dissertation. Additional information about
each of these topics is presented in subsequent chapters. The intent here is to help contextualize
the forthcoming more focused discussion.

2.1

Environmental Background
The environmental history of the UP begins with the retreat of the Wisconsinan ice sheet

and the deglaciation of the region around 11,500 years before present (BP) (Dorr and Eschmann
1986). This period saw the origins of the modern Great Lakes, although in a significantly
different configuration, which were fed by glacial melt water (Larson 1999). Great Lakes levels
exhibited great variation over the next seven thousand years (through 4500 BP), as did the
climate and vegetation (Futyma 1982; Kapp 1999; Larson 1999; Lovis et al. 2012). Forest
communities similar to those of the present were established in the Upper Great Lakes region by
3000 years ago and modern levels of the Upper Great Lakes had been generally achieved by
2000 years ago. Therefore, modern lake elevations and configurations encompass much of the
Woodland period (from ca. AD 0 to AD 1600) in the region (Anderton 1993; Brugam et al.
1997; Winkler et al 1986).

13

This does not mean that the UP’s environment has been static for the last 2000 years, but
rather that the regional environment was broadly resilient (sensu Resilience Theory), subject to
localized changes resulting from disturbance regimes and climatic factors (see Baedke and
Thompson 2000; Lovis et al. 2012; Ritchie 1986; Zhang et al. 1999). For example, the water
levels of the Great Lakes regularly fluctuated by up to a meter on a fairly regular short term cycle
and on a wider scale at a longer interval (Baedke and Thompson 2000). Additionally,
temperature and relative moisture varied across the region and there were periods of dune
activation and stabilization along the Great Lakes shores (Arbogast 2009; Bernabo 1981; Booth
et al. 2004; Delcourt et al. 2002; Futyma 1982; Lovis et al. 2012). An attempt at illustrating
these trends is presented in Figure 2.
The lake level and dune formation cycles may have been coupled to dynamic vegetation
change in the littoral zones along the Great Lakes. Temperature and moisture change would also
have the potential to alter vegetative patterns across the UP in general. The significant changes
in forest composition over the past 150 years, namely the reduction of tamarack, hemlock and
white pine and concurrent increase in red maple, sugar maple, and red oak, can serve as a useful
example (Comer et al. 1995; Leahy and Pregitzer 2003; Nowacki and Abrams 2008; Nowacki et
al. 1990; Price 1994; Van Deelen et al. 1996; Whitney 1986; Zhang et al. 2000). This
reconfiguration of northern forests is the result of intensive logging and subsequent fires in the
second half of the nineteenth century and most of the twentieth century (see Dey 2002). Oak
benefited from the removal of canopy (it is not a shade tolerant species) as well as burning of
subsequent understory (see Abrams 1992; Crow 1988), and most varieties of maple are
aggressive colonizers in openings. Maple and oak also benefited from increased deer

14

Figure 2: Timeline showing selected cultural and physical environmental variables relating
to the EUP. The data is subdivided by three regions: The Straits of Mackinac (Straits) and
Northern Lake Huron; Bay de Noc and Northern Lake Michigan; and the South Shore of Lake
Superior (see Figure 1.0-1). Physical environmental variables are presented in terms of higher or
lower than present day (in relation to Great Lakes level and relative temperature) or
active/inactive (coastal dunes). Cultural variables are presented as present/absent based on direct
or indirect archaeological evidence. The references for the data presented in the figure are as
follows: 1(Drake and Dunham 2004; Dunham and Hambacher 2007); 2(B. A. Smith 2004; also
see Cleland 1982; Martin 1985]); 3(Loope et al. 2004); 4(Dunham and Hambacher 2002);
5
(Buckmaster 2004); 6(McPherron 1967); 7(O’Shea 2003); 8(Lovis et al. 2012); 9(Baedke and
Thompson 2000); 10(Bernabo 1981).
populations, a species that also benefitted as a result of land clearing through logging. Deer
browsing on sensitive species such as hemlock, yew, and cedar furthered their decline, while
species less favored by browsing deer increased (Alverson et al. 1988; Van Deelan et al. 1996).

15

There is also variation in the ecosystems present across the geographic extent of the
eastern UP (see Albert 1995). An ecosystem is a structure in which a community of living
organisms (e.g., plants, animals [biotic community]) interacts with each other as well as the
nonliving component of their environment (e.g., air, water, soil [abiotic components]). The
interactions of these variables help to sustain one another in regular patterns. For example,
maize is an exotic plant that was brought into the Upper Midwest by Native American people as
a food resource (Hart and Lovis 2013). Maize requires a series of ecological variables including
soil, rainfall, and frost free days to produce a reliable subsistence crop (Demeritt 1991; Hart and
Lovis 2013; O’Shea 2003; Yarnell 1964). Modern maps showing growing season (frost free
days) indicate that areas near the shorelines of Lakes Michigan and Huron have the required
growing season, but most of the eastern UP does not (Eichenlaub et al. 1990; see also O’Shea
2003). Thus, variation in ecosystem characteristics can play a strong role in what plants (such as
maize or red oak) or animals (such as deer) might be present.
The eastern UP is bounded on the north by Lake Superior and on the south by Lakes
Michigan and Huron. There are numerous river drainage ways in the eastern UP (Figure 3).
Most of these drain either north or south with important exceptions like the Tahquamenon River
which flows generally east. Figure 3 illustrates that there are more drainages with larger
watersheds that flow south into Lakes Michigan and Huron than into Lake Superior. The largest
drainages are the Manistique (4564 km² [1762 mi²]) which flows generally south into Lake
Michigan and the Tahquamenon (2549 km² [984 mi²]) which flows into Lake Superior. Each of
these rivers drain the expansive Seney wetlands in the interior of the eastern UP (Appendix 2A).
The Manistique watershed is also noteworthy in that the two largest inland lakes in the eastern

16

Figure 3: River Drainages in the Eastern Upper Peninsula.
UP are part of the watershed. Manistique Lake (10,130 acres) forms the headwaters of the
Manistique River and Indian Lake (8,000 acres) is at the confluence of the Manistique River and
its major Tributary, the Indian River. The Indian River drainage, which forms the western
portion of the Manistique watershed, includes a watershed of approximately 453 km² (174.9 mi²)
on its own.

2.2

Cultural Background

2.1.1 General Overview of the Woodland Period in the Eastern UP
Archaeologists have developed a broad chronological and cultural classificatory scheme
to organize and describe the prehistory of eastern North America. The following subdivisions are

17

broadly applicable, although there is regional variation both in chronology and culture (c.f.,
Fitting 1975a; Mason 1981; Snow 1976); Paleoindian (12,000 BC to 8,000 BC), Archaic (8,000
BC to 1000 BC), and Woodland (1000 BC to AD 1600). Throughout much of the Midwest the
Woodland period is divided into three parts; Early Woodland (1000 BC to 200 BC), Middle
Woodland (200 BC to AD 600), and Late Woodland (AD 600 to AD 1600). The standard
hallmark for the beginning of the Woodland period is the introduction or inception of ceramics.
In northern Michigan, including the eastern UP, the Archaic period persists about 1000 years
longer than in more southerly parts of the Midwest (see Brose and Hambacher 1999; Robertson
et al. 1999). Ceramics do not appear in the eastern UP until what is considered the Middle
Woodland time period in the Midwest, and those are wares representing the Laurel and North
Bay traditions (Brose 1970; Brose and Hambacher 1999; Fitting 1975a; Janzen 1968; Mason
1981). The temporal lag in ceramic introduction across the region has prompted some
researchers to substitute the term “Initial Woodland” for this period; here, the standard
Midwestern convention of Middle Woodland will be employed.
The Middle Woodland period (AD 1 to AD 500) in the Upper Great Lakes represents the
first widespread introduction of ceramics in the Upper Peninsula. In general, Middle Woodland
sites in the Upper Peninsula may represent Lake Forest (Fitting 1975a) or Northern Tier (Mason
1966) adaptations sharing material culture affinities with Laurel sites to the south, north, and
west (Janzen 1968). Settlement and subsistence patterns suggest seasonal fishing, collecting, and
hunting, with an increasing emphasis on exploitation of aquatic resources (Brose and Hambacher
1999; Cleland 1982). Sites such as Summer Island (Brose 1970), Winter (Richner 1973), and
Naomikong Point (Janzen 1968) are interpreted as Middle Woodland warm season fishing
villages. In the St. Ignace area a number of sites having Middle or Initial Woodland components

18

are thought to have been satellite summer or winter camps (Fitting and Clarke 1974; Fitting [ed]
1974). Spider Cave on the Garden Peninsula may have been a ritual locale (Cleland and Peske
1968).
Other recognized Middle Woodland sites within the region include the Nina Site
(Dunham and Hambacher 2002); the stratified Bark Dock site in Chippewa County (Dunham and
Hambacher 2007); the Gooseneck Lake IV site (Franzen 1987), the Indian River site (Franzen
1987), the multicomponent Williams Landing locale (Dunham and Branstner 1995), and the
Carp River site (Dunham et al. 1993). The Carp River site has been interpreted as a transitional
Middle-Late Woodland fishing encampment at the mouth of the Carp River near the Mackinac
Straits (Dunham et al. 1993). Faunal evidence from the Carp River site suggests an emphasis on
spring-spawning species (e.g., sturgeon walleye, etc.), although fall spawning species (e.g.,
whitefish and lake trout) are also represented. The location of the site is also consistent with
transitional

Middle-Late

Woodland

fishing

locales

based

on

current

regional

settlement/subsistence models (Cleland 1982; S. Martin 1989; B. A. Smith 2004) as the mouth of
the Carp River would provide excellent access to both river-spawning spring species and deep
water-spawning fall species.
The Late Woodland period (AD 500 to AD 1600) is perhaps the best-documented
cultural period in the northern Great Lakes and UP. The best known LW site in the eastern UP is
the Juntunen site (20MK1; McPherron 1967). This stratified site is located on an island in the
Straits of Mackinac and has provided the basic chronology for the LW in the region and beyond:
Mackinac Phase, AD 800 to AD 1000; Bois Blanc Phase, ca. AD 1000 - AD 1200; and Juntunen
Phase, AD 1200 - AD 1500. It also has produced a large artifact assemblage with the remains of
more than 1600 ceramic vessels and 400 formal stone tools, a large faunal assemblage, and

19

features, including structural remains. Not only is the Juntunen site the best known in the region,
it is the basis for much of the Inland Shore Fishery model (Cleland 1966; 1982).
Four sites in the Bay de Noc region, at the northern end of Lake Michigan, provide
interesting information concerning the settlement and subsistence of the region. There are two
components of the Summer Island site that fall within the Late Woodland period (Brose 1970).
The first can be generally classified as Oneota (Upper Mississippian) and was occupied ca.
thirteenth and fourteenth century AD and included 16 ceramic vessels as well as ground stone
and chipped stone tools. This component also included floral and faunal materials as well as
evidence or refuse and storage pits (Brose 1970). The later component was occupied at the end
of the Late Woodland period (ca. sixteenth century) and into the early historic period (Brose
1970). The late occupation is also evidenced by a small assemblage of European produced trade
goods. This late component also included Oneota ceramics (13 vessels) and produced ground
stone and chipped stone tools.

Importantly, this component also produced evidence for a

structure as well as squash seeds (Brose 1970).
The Ogontz Bay site, situated on the shore of Big Bay De Noc, has produced Sand Point
ceramics (cf., Dorothy 1980), a variety of stone tools, and a very diverse faunal assemblage,
including fish, mammals, and birds (Anderton et al. 1991). The faunal assemblage suggests a
focus on spring-spawning fish species (especially walleye and bass). Other intriguing faunal data
include a cache of black bear mandibles, human-modified painted turtle shell, beaver incisors,
and a bald eagle talon. These finds may indicate specialized resource procurement and may,
particularly in reference to the eagle claw, reflect ideological behaviors.
Archaeological testing indicated that the Bar Lake site, an inland lake site, was used as a
hunting and fishing camp between AD 1100 and AD 1600 (Dunham and Hambacher 2002:71-

20

107). The remains of at least nine fragmentary Oneota vessels and a variety of stone tools were
found, including projectile points and tools used for cutting, scraping, and grinding. A variety of
animal bones was also found. The artifacts and animal bone indicate moose and beaver were
hunted at the site. The 10 Mile Rapids site, situated inland on the Sturgeon River, has produced
cordmarked ceramics, a triangular point, and a slate elbow pipe, as well as a faunal assemblage
consisting primarily of sturgeon and beaver bone (Rutter et al. 1984). Both the site’s location at a
rapids and the recovered faunal assemblage indicate that it was a spring encampment focusing on
spawning sturgeon.
In the Lake Superior Basin, few Late Woodland sites have been formally excavated.
Naomikong Point contains a Late Woodland component (Janzen 1968). Ceramic styles across
much of the eastern part of the Superior basin indicate an affinity towards the south and east,
including Juntunen wares and Iroquoian motifs. In the central and western UP, along Lake
Superior, the ceramic wares are more westerly in association. For example, the Sand Point site in
Baraga County, which has produced Blackduck and Oneota-like ceramic assemblages, appears to
represent a village and funerary mound (Cremin 1980). Similarly, the Gete Odena site at the
Williams Landing locale on Grand Island has produced Madison points and Sand Point-like
ceramics (Dunham and Branstner 1995; Robinson et al. 1991). The 1994 excavations at the latter
site identified subsurface features and a possible former living floor associated with the Late
Woodland occupation of the site (Dunham and Branstner 1995). One feature, a small pit,
contained a miniature ceramic vessel as well as the carbonized remains of cherry and acorn.
These excavations also revealed three grooved sandstone net sinkers, reflecting the use of the
adjoining bay for deep water fishing. Similar fishing weights, as well as Oneota and Sand Point

21

ceramics, have been recovered from the Bark Dock site in Chippewa County (Dunham and
Hambacher 2007).
Standing models of Woodland settlement and subsistence in the eastern UP relate that
Middle Woodland people (AD 1 – AD 600) were likely more residentially mobile than their Late
Woodland descendants and had a broader diet breadth (Brose and Hambacher 1999; Cleland
1983; S. Martin 1985). As Late Woodland peoples (AD 600 – AD 1600) became more reliant on
aquatic resources, specifically the integration of and intensification on the fall fishery, they
became more territorially constrained and more socially integrated/organized (Cleland 1982;
1992a; McHale-Milner 1991; B. A. Smith 2004; see also Binford 2001). Change towards more
circumscribed territories, more focused subsistence, and more complex social organization can
be seen throughout the Midwest during the Late Woodland period (Mason 1981; McElrath et al.
2000; McHale-Milner 1991; Schroeder 2004; Seeman and Dancey 2000). In the eastern UP, the
focus of this shift was the exploitation of the fall, deep-water fishery (Cleland 1982; B. A. Smith
2004; for an alternative discussion, see S. Martin 1989). This provided a seasonally abundant,
storable food resource to offset the diminished food potential of the eastern UP over the course
of the winter.

2.2.2 The Late Woodland Period
A number of cultural changes occur during the course of the Woodland period in the
Upper Great Lakes region that warrant summarization. Residential sites become larger due to
both reoccupation cycles and the size of residential groups, and technological innovations such
as ceramics and new tools for fishing appear. Subsistence strategies consist of seasonal fishing,
collecting, and hunting, with an increasing emphasis on aquatic resources. The prevailing model

22

focuses on technological and social changes resulting from the exploitation of spring and fall
spawning fish (Cleland 1982; B. A. Smith 2004). In simple terms, Middle Woodland (AD 1 to
AD 600; also known as the Initial Woodland in the Upper Great Lakes region) fishing focused
on spring-spawning species such as sturgeon and sucker. In the Late Woodland (sometimes also
known as Terminal Woodland [AD 600 to AD 1600]) period people continued to follow a
seasonal settlement and subsistence pattern, in many ways quite similar to that of the Middle
Woodland, with the critical addition of deep water, fall-spawning fish such as whitefish and lake
trout.
Cleland (1982) has constructed a technoeconomic model in which the development of gill
net fishery technology represents the cornerstone of a series of changes in resource use and site
placement as well as social transformations in the Late Woodland period. He hypothesizes that
the adoption of deep water gill-nets to catch surpluses of fall-spawning fish, and the utilization of
an effective storage technology increased the availability of food during the cold season, which
in turn allowed for population increases, the development of larger settlements of increasing
residential duration, and cooperation among social groups (inter- and intra-group). Other Late
Woodland people in the region were focusing on maize and wild rice which, like fall-spawning
fish, could be stored against winter shortfalls (Brashler et al. 2000; Cleland 1983; Gibbon and
Caine 1980; Moffat and Arzigian 2000; Vennum 1988). Some have suggested the presence of an
interdependent, or symbiotic, relationship between specialized interior hunter/foragers and
coastal maize horticulturalist in parts of northern lower Michigan (O’Shea 2003; see also Howey
2006).
Cleland’s (1982) model is largely based on evidence derived from archaeological sites
situated on the Great Lakes shorelines, as well as ethnographic and ethnohistoric accounts which
23

discuss the Great Lakes fishery as part of a seasonal settlement and subsistence round (see also
Cleland 1983, 1992a). It is a coastal oriented model, but includes mechanisms that would allow
people to use the interior. For example, it is commonly assumed that aside from logistic hunting
and collecting forays, the interior was seasonally occupied by smaller groups of people who
dispersed there during the winter (see Fitting and Cleland 1969; Quimby 1962; see also Holman
and Lovis for a discussion of the ethnohistoric basis of the model). In basic terms, the
subsistence round is centered on two axes - spring and fall fishing. The underlying logic is that
people came together to harvest seasonally dense resources (spring and fall spawning fish) and
dispersed when resources were more scarce such as in the cold season, or were more broadly
distributed across the landscape (as in the warm season).
The shift towards the fall fishery was the result of new technologies and social practices –
specifically deep-water gill nets, storage technology (drying and freezing), the development of
larger settlements of increasing duration of occupation (permanence), and cooperation among
social groups (intra- and inter-group) (Cleland 1982). The combination of gill nets, social
cooperation, and storage are critical to the success of this process. In a Northwest Coast example
Schalk (1977:240) highlights the difficulties in processing and storing anadromous fish referring
to a “bottleneck effect of having to harvest and process at the same time” due to rapid rates of
spoilage (emphasis in the original). The effort requires an increased level of social organization
and this leads to a combination of practical and social storage (see Binford 2001; Ingold 1983;
O’Shea 1981). In other words, the intensive group processing of fish for storage is carried out, in
part, with the understanding that it will be available for future use by the group engaged in its
processing (see Cleland 1982).

24

The Cleland model of the Late Woodland use of the Great Lakes deep fresh water fishery
is compelling, but there are several constraints that are not fully explored. A primary constraint is
related to the season of fall fishing. The harvest of fall spawning fish took place on the Upper
Great Lakes primarily between mid-November and late December. Cleland’s model requires that
a portion of the fish caught and processed be dried and frozen for storage. Therefore, air
temperature is a factor in storage preparation. If the temperature is too warm and the surplus fish
cannot be frozen, it may be subject to spoilage. If the temperature is too cold, the lake may begin
to freeze limiting or preventing access to spawning beds until the ice was thick enough to support
net disbursement through the ice. Other factors, such as rough seas could also lead to a failure to
collect fish. In this instance, the ability to generate a surplus of food and store it prior to winter is
critical to nutritional success and survival over the winter.
One solution to the question of subsistence risk has been offered by Holman and Lovis
(2008) who see a relationship between the highly flexible mobility strategies of Upper Great
Lakes peoples along with integrating social mechanisms as a buffer for environmental
variability. The type of mobility practiced as part of the so-called Chippewa and Ottawa patterns
(see Fitting and Cleland 1969) appear to have a relationship with their respective environments.
The cooperation between these groups facilitated movement into adjoining territories in times of
subsistence stress as well as the sharing of territories/resources at other times (see also Holman
and Kingsley 1996; McHale-Milner 1991). The combination of highly flexible mobility and
social cooperation may mitigate some of the risk posed by environmental variability in the
eastern UP, but likely not all of the risk.

25

An additional risk buffering mechanism is the integration of alternate, highly productive
food resources into the system (O’Shea 1989; see also Gallagher and Arzigian 1994; Parker
1996). Cleland’s (1982) model would similarly benefit from the presence of other highly
productive, storable resources that could serve to buffer against a poor fish harvests. As noted
above, maize and wild rice filled this niche in the regions immediately south and west of the
eastern UP (Brashler et al. 2000; Cleland 1983; Moffat and Arzigian 2000; O’Shea 1989; Parker
1996). High carbohydrate plant resources such as wild rice and/or maize would complement the
fish diet of Late Woodland peoples and provide a potential buffer to environmental risk in the
eastern UP (Cordain et al. 2000; Speth 1991).
Pilot studies addressing broader settlement and subsistence patterns in the eastern UP
suggest a shift in site locations from the Middle Woodland to Late Woodland periods (Dunham
2002; see also Buckmaster 1979; Drake and Dunham 2004). These findings reveal that access to
deep water settings on the Great Lakes shorelines as well as site locations on the interior are of
greater importance to Late Woodland people than they were in the Middle Woodland. This
apparent shift is highlighted by an overall increase in the number of Late Woodland sites as well
as a shift in site locations. The shift in Late Woodland coastal sites towards deep water locales is
likely the result of the development of the fall fishery, whereas the shift towards greater use of
interior settings may reflect changes in mobility strategy and/or landscape positioning keyed
towards the adoption of new resources or intensification on existing resources. While the timing
of the shift is not known, largely because of the low number of chronometrically well dated sites
in the eastern UP, the data supports such a trend.

26

If it can be assumed based on Cleland’s (1982; see also S. Martin 1985; B. A. Smith
2004) model that access to spring and fall spawning beds is an important factor in the occupation
of coastal zones, what might the expansive interior areas of the eastern UP, largely ignored by
these models (see Dunham 2002; Franzen 1986; 1987; Holman 1978), have to offer Late
Woodland peoples?

Holman (1978) observed that early Late Woodland (Mackinac Phase)

coastal sites in northern lower Michigan, immediately south of the current area of interest, had
high late fall resource potential because of the fall fishery, but relatively low winter resource
potential. Interior sites, on the other hand, were situated in areas of high(er) winter resource
potential. She concluded that early Late Woodland people chose to live in the interior over the
winter to make use of the higher winter resource potential (Holman 1978:53: Lovis 2008). An
important exception to this general pattern is the Juntunen site which had high resource potential
year round (Holman 1978). While coastal and interior settings each had relatively high warm
season resource potential, people may have more commonly lived in coastal areas for other
social and economic reasons (Holman 1978; see also Holman and Lovis 2008).
Through recent pilot studies, Dunham (2008, 2009) has identified acorns and wild rice as
likely resources used by Late Woodland peoples in the interior of the eastern UP. Likewise,
O’Shea (2003) has suggested that microclimates associated with the ameliorating affects of Great
Lakes coastal zones may have permitted localized maize horticulture. Wild rice and maize are
not typically associated with eastern UP Late Woodland foodways largely due to perceived
environmental constraints and acorns have often been overlooked because of their perceived low
density and high processing costs (Cleland 1983; Dunham 2009; Yarnell 1964). The studies by
Dunham (2008; 2009) and O’Shea (2003) demonstrate that wild rice, maize, and acorns may be
abundant in specific niches and should be considered. Each of these resources is relatively

27

predictable, can be stored, and compare favorably with one another in general nutritional
characteristics (Dunham 2009; Kuhnlien and Turner 1991). They are also significant sources of
carbohydrates which would complement fall spawning fish, and, importantly, the procurement of
these plant resources would not conflict with the scheduling of fall fishing. Another important
consideration for each of these resources, including fall spawning fish, is that they each have
predictable spatial, environmental, and temporal (seasonal) constraints.
While a significant amount of archaeological research has been conducted in the eastern
UP, a correspondingly proportionate data set relating directly to Late Woodland subsistence has
not been generated (e.g., floral and faunal remains [see Chapter 3.5]). Two factors are largely
responsible for this: 1) much of the data is the result of archaeological survey and limited test
excavation (see Anderton et al. 1991; Dunham and Branstner 1998; Dunham and Hambacher
2002; Dunham et al. 2010; Franzen 1987); and 2) preservation of floral and faunal remains is
typically poor in the region due to acidic soils and slow rates of soil development. There are
exceptions to these trends. Sites that have had larger scale excavation and include floral and
faunal remains (e.g., the Juntunen site [McPherron 1967; see also Cleland 1966; Yarnell 1964]).
However in many cases the archaeological data are limited to lithic and ceramic assemblages.
Thus, there are archaeological site locations without much corresponding information on the
resources that were being used at these sites.
Some have approached the incongruity between Late Woodland site locations and
subsistence in northern Michigan through models derived from environmental analyses coupled
with ethnographic and ethnohistoric sources (see Cleland 1966, 1982, 1992a, 1992b; Dunham
2000a; Fitting and Cleland 1969; Franzen 1986; Holman 1978; B. A. Smith 1996; Yarnell 1964).
The working assumption is that Native Americans in the historic period were interacting with

28

environments broadly similar to those that Late Woodland people operated in and they were
likely using the same subsistence resources (Franzen 1986; Holman 1978; Holman and Krist
2001; Holman and Lovis 2008; S. Martin 1985). The known pattern of seasonal and
interseasonal mobility also demonstrates that Late Woodland people had the opportunity to
access and use a wide variety of plant and animal resources across the landscape in addition to
spring and fall spawning fish.
While the proposed study is geographically specific to the eastern UP, it is understood
that more than one settlement and subsistence strategy may be present both spatially as well as
temporally (inter-annually, intra-annually, etc.) in this region. Archaeological data suggest that
there may be multiple cultural traditions/ethnic identities in the eastern UP (Figure 2). In the
early Late Woodland period (ca. AD 600 to AD 1000) the ceramic assemblages are culturally
more western oriented (Brose 1970; Dorothy 1980; McPherron 1967). In the Straits of Mackinac
and northern Lake Huron basin ceramic trends follow the sequence derived from the Juntunen
site where the more westerly influences of the early Late Woodland period are subsumed by
more eastern, Ontario Iroquoian-like, influence during the later Late Woodland (McPherron
1967). The presence of Oneota wares in the Bay de Noc region suggests that this region is part of
a different cultural system than the Straits by the late Late Woodland period (Brose 1970;
Buckmaster 1979; Dunham and Hambacher 2002). The late Late Woodland assemblages along
the south shore of Lake Superior include Sand Point, Oneota, and/or Juntunen related wares
linking them to the east, west, and south (Dorothy 1980; Drake and Dunham 2004; Dunham and
Branstner 1995; Dunham and Hambacher 2007; Dunham et al. 2010). Such patterns may reflect
indicators of identity (e.g., tribal identity) and territoriality (see O’Shea and Mc-Hale-Milner
2002).

29

Other potential indications of cultural differences may be reflected in the relative size of
sites, with potentially more nucleated sites (i.e. more constrained and denser occupations
indicating focused spatial use) being located in the Straits of Mackinac and Bay De Noc regions
and more dispersed settlement locales being utilized along the south shore of Lake Superior
(Brose 1970; Drake and Dunham 2004; Dunham 2002; Lovis and Holman 1976; McPherron
1967). Further, the use of interior locations is significantly higher in the Bay de Noc region than
either the Straits of Mackinac or Lake Superior regions (Dunham 2002).
Likewise, subsistence technology varied through the Late Woodland. A recent
reevaluation of the data relevant to the fall fishery found that the increased reliance on fall
spawning whitefish and lake trout began to appear around AD 800 at the Juntunen site, but after
AD 1100 in northern Lake Michigan basin and as late as AD 1400 in the rest of the Upper Great
Lakes region (B. A. Smith 2004). The geographically closest direct evidence for maize
horticulture in the Late Woodland period comes from the Menominee River in the southern
portion of the Bay de Noc region with directly dated maize cupules and a ridged field complex
from about AD 1400 (Buckmaster 2004). These examples further illustrate the need to reassess
our understanding of Late Woodland settlement and subsistence strategies.
In addition to archaeological data, there is a significant body of paleoenvironmental data
for the eastern UP that includes studies of lake level variability and dune formation over the past
two millennia (Anderton 1993; Baedke and Thompson 2000; Loope et al. 2004; Lovis et al.
2012); paleoclimatic data reflecting temperature variation (e.g., Medieval Climatic Optimum and
Little Ice Age) and precipitation (Bernabo 1981; Booth et al. 2004; Delcourt et al. 2002); and
paleoenvironmental data including long term pollen trends and pre-European settlement land

30

cover (see Albert and Comer 2008; Bourdo 1954; Brubaker 1975; Davis et al. 2000; Delcourt
and Delcourt 1996; Futyma 1982; Price 1994; Woods and Davis 1989; Zhang et al. 2000).
Additionally, a recent study has found that taphonomic processes may have led to better
preservation of archaeological sites in certain coastal zones around AD 1000 (Lovis, Monaghan
et al. 2012). If this is the case, how might these better preserved sites enhance and/or skew our
understanding of coastal settlement and subsistence dynamics in the middle part of the Late
Woodland? Studies such as these, that explore the dynamic nature of the environmental context
of the eastern UP during the Late Woodland period, are critical to this research.
The questions of whether, why, and how cultural and/or environmental changes occur fit
well within the application framework of Resilience Theory. Multiple adaptive cycles seem to
converge over the first half of the Late Woodland period (ca. AD 600 to AD 1100) (see Chapter
1.3; see also Figure 2). These include technological, social, and organizational changes as well as
climatic and associated environmental changes each of which have the potential to affect
multiple variables. Each of these changes may represent the completion of an adaptive cycle
from the perspective of Resilience Theory (Redman 2005; Redman and Kinzig 2003; Walker et
al. 2006). Resilience Theory, therefore, will serve as a useful heuristic framework for this
project.

2.3

Background Discussion
The purpose of this chapter was to provide an overview of the environmental background

and cultural setting for the research conducted in this dissertation. To summarize the cultural
background, the existing models of Woodland settlement and subsistence in the eastern UP relate
that Middle Woodland peoples (AD 1 - AD 600) were likely more residentially mobile than their

31

Late Woodland descendants and had a broader diet breadth (Brashler et al. 2000; Brose and
Hambacher 1999; Cleland 1983; S. Martin 1985). As Late Woodland peoples (AD 600 – AD
1600) expanded their technological capabilities and became more reliant on seasonally abundant
aquatic resources, specifically the integration of and intensification of the fall fishery into the
subsistence round, they became more territorially constrained, more subsistence focused, and
more socially integrated/organized at different scales (Cleland 1982, 1992a; Holman and Lovis
2008; McHale-Milner 1991, 1998; O’Shea and McHale-Milner 2002). It is unclear precisely
when these changes occurred, but they appear to have taken place between AD 900 and AD
1100. The changes towards more circumscribed territories, more focused subsistence, and
greater social organization can be seen throughout the Midwest, and the Eastern Woodlands,
during the Late Woodland period (Brashler et al. 2000; Mason 1981; McElrath et al. 2000;
McHale-Milner 1991; Seeman and Dancey 2000).
This dissertation revisits the topic of settlement patterns and subsistence practices in the
Late Woodland period of the eastern UP and evaluated it against the body of data generated over
the past 25 years. New data suggests that the dominant settlement and subsistence model is
incomplete, lacks explanatory value, and requires revision. This study tests the hypothesis that a
suite of potential resources was both present and utilized, allowing for a more flexible set of
strategies, i.e. it is not based on a single primary resource (the fall fishery). Archaeological
evidence, ethnographic data, and pilot study results reveal that acorns, maize, and wild rice are
likely resources to be incorporated into such a strategy; all can be harvested and stored in the late
summer or fall as a buffer against a poor fish harvest. Each, however, also has spatial,
environmental, and temporal constraints with implications bearing on archaeological site
locations as well as the evidence from the sites themselves.

32

The next two chapters explore these topics. Chapter 3 will explore the material remains
recovered from Late Woodland sites in an attempt to glean if these sites were used for different
purposes. Chapter 4 will examine the spatial distribution of the sites to determine if the sites
were spatially proximate to particular resources. The Late Woodland period can be easily
divided into two sub-periods as outlined above: an early Late Woodland (ca. AD 600 to AD
1000) and a late Late Woodland (ca. AD 1000 to AD 1600). The analyses carried out in
Chapters 3 and 4 will consider the temporal variable (early and late Late Woodland) as well as
broader geographic distribution in regard to settlement and subsistence patterns. Environmental
data will be integrated into the discussion as yet another variable in understanding these trends.

33

3.0

ARCHAEOLOGICAL ASSEMBLAGES
The following chapter will provide a review and synthesis of the available archaeological

data for Late Woodland sites in the eastern UP. The review relied on previously completed
studies and includes published and unpublished sources including technical reports comprising
the so-called “gray literature” (see Seymour 2010). As noted in Chapter 1, the eastern UP is
defined as Alger, Chippewa, Delta, Luce, Mackinac, and Schoolcraft Counties (see Figure 1).
That definition is, in part, an outcome of the organization of the archaeological site files at the
Michigan State Historic Preservation Office (SHPO).
There are two LW data sets that are used in this study. The previously cited pilot studies
(particularly Dunham 2002; 2009), identified some 76 archaeological sites in the eastern UP with
at least one LW component. The current study has expanded this number to include 81
archaeological sites that include LW components that are listed in the archaeological site atlases
and files at the SHPO and the Hiawatha National Forest (HNF) (Figure 4; Appendix B). The
first data set includes 48 archaeological sites with Late Woodland components that have been
discovered through archaeological survey on the HNF. The second, including 33 archaeological
sites with Late Woodland components, is the balance of the 81 sites identified. The HNF set
were discovered through relatively consistent survey methods (see Franzen 1986; Anderton et al.
1991; Dunham and Branstner 1998; Dunham et al. 2010; Rutter and Weir 1985), whereas the
remaining sites have been identified in a variety of ways including formal survey, informant
sources, etc.
The available reporting was reviewed for each of the LW sites in the eastern UP. The
archaeological site data were summarized in tabular form to establish a format to compare the

34

Figure 4: Locations of the 81 Late Woodland Archaeological Sites.
characteristics of each assemblage (Appendices B - I). This information includes, where
possible, the site designation (number), site location (UTM coordinates), estimated size of the
site (square meters [m²]), whether the site is known from survey or excavation, the number of
components (single or multiple), and how much of the site has been excavated (m²).
Information was also collected on the LW assemblages at each site including: the chipped
stone assemblage; ground stone tools; fire-cracked rock (FCR); ceramics; floral remains; faunal
remains; features (Appendix B); and chronometric dates (Appendix B). The way that the
individual assemblages were analyzed, coded, and recorded varied, so the terminology and
techniques were not consistent. The analyses as presented by the original researchers were
maintained and no new analyses of the artifact assemblages were conducted.

35

In some cases it was possible to reorganize data presented in these reports to better
address the needs of this study. An attempt was made to present only the LW component of
multicomponent sites, such as at the Bark Dock site (20CH95), which includes Middle
Woodland and LW components (Dunham and Hambacher 2007) and Gete Odena site
(20AR348) which includes Archaic, Middle Woodland and Historic components in addition to a
LW component (Dunham and Branstner 1995; Robinson et al 1991; Skibo et al. 2004)
(Appendix C). Likewise, when multiple LW components were present on a site, an attempt was
made to differentiate them as separate components as well (e.g., the Juntunen and Summer Island
sites [Brose 1970; McPherron 1967]). This was not always possible and depended on the
reporting of the site.

As a result, a small number of sites (n=7) were characterized as LW

locales based on the recovery of LW artifacts, but where the LW component could not be
differentiated from other components on the site (e.g., the palimpsest effect [see Bailey 2007])
(see Appendices B and C). In at least one case, 20AR338, the results of the analyses are ongoing
and specific assemblage composition pertaining to the LW component of the site is not yet
available (see Drake et al. 2009; Dunham et al. 1997; Skibo et al. 2009).
The data from the LW components has the potential to more fully elucidate subsistence
resources used, site function, seasonality of use, and social relationships among other things.
The quality of data ranges significantly from site to site in this sample. The information
available for each site is variable as a result of the different level of investigation at each site.
The data are derived from archaeological survey, small excavations, and larger excavations. As
a result, it is not possible to classify the function, seasonality, or precise age of many of the sites.

36

3.1

Chipped and Ground Stone Artifacts

3.1.1 Chipped Stone Tools
For chipped stone artifacts, information was recorded for formal tools (such as projectile
points, bifaces, and scrapers) and informal tools (retouched and edge damaged flakes), as well as
debitage (flakes) (Appendix D). Where possible, the formal tools were typed following their
original recordation. Debitage was minimally recorded by count. When reduction stage and raw
material type were identified, this information was also recorded.
Nine hundred and seventy nine of the chipped stone artifacts are classified as formal tools
and 919 as expedient tools. Expedient tools account for 48.4 percent of all the chipped stone
tools. The remainder of this discussion will focus on formal chipped stone tools. Formal
chipped stone tools were recovered from 54 LW sites or components (Appendix D). The
Juntunen site (20MK1) produced the highest number of formal chipped stone tools (n=440), but
only 161 could be securely placed in the three primary components (The Mackinaw, Bois Blanc,
and Juntunen Phases). The Juntunen and Mackinac Phases of the Juntunen site (20MK1), along
with the proto-historic component at the Summer Island site (20DE4) produced the highest
number of chipped stone tools (n=70, 63, and 61, respectively) and fifteen LW sites produced
only a single chipped stone tool.
The formal tools fall into five primary categories: projectile points; scrapers (end, side,
and unspecified); bifaces; drills; and knives. Bifaces are characterized by the presence of retouch
on both faces of the tool, but cannot be classified as another category, such as projectile points.
Projectile points are the best represented formal tool type, appearing on 39 LW sites. Scrapers
were recovered from 37 sites, but were not uniformly classified in the various reports. End
scrapers were recorded on 31 sites and side scrapers on 11. Unspecified scrapers were recorded

37

from six components including all the scrapers (n=245) from the Juntunen site. Thirty sites
included bifaces. Four sites produced drills and one site a knife. The knife might be better
classified as a biface, but the reporting was not clear enough to make this consclusion.
The various tool classifications can provide insight into tool function, although a given
tool may be used for a wider variety of ways than these baseline attributions (Kooyman 2000).
Projectile points are most commonly associated with hunting activity, so their presence on a site
is indicative of hunting. Scrapers were mostly used for hide preparation, but may have had
numerous other uses during their use life prior to discard. The presence of scrapers on a site
reflects the processing of animals. Drills were used to perforate things and are usually associated
with perforating harder objects. Bifaces may encompass cutting tools, chopping tools, or
unfinished tools.
In many contexts, projectile points are a useful temporal indicator for dating
archaeological sites or components of archaeological sites (see Justice 1987). Unfortunately, this
is not always the case in the eastern UP. Triangular projectile points, sometimes broadly referred
to as Madison points, are diagnostic of the Late Woodland period in the region and their
presence on a given site or component is sufficient to define LW activity (Brose 1970; Fitting
1975a; Janzen 1968; McPherron 1967). Another form that is diagnostic to the LW is small flake
points known as Juntunen points (McPherron 1967). Other points found in LW contexts are
stemmed, side notched, or corner notched, characteristics which also appear in earlier Middle
Woodland or Archaic assemblages (Brose 1970; Cleland and Peske 1968; Janzen 1968;
McPherron 1967). For this discussion, points were classified as single tool type and were only
used to designate the relative age of a site (LW) if the point(s) were triangular or so-called
Juntunen flake points.

38

3.1.2 Chipped Stone Debitage
Fifty six of the LW components considered in this study produced debitage that are
considered in this study (Appendix D). Four additional sites produced no debitage. The
debitage assemblage of 21 of the LW sites cannot be formally discussed for a variety of reasons.
The LW components that produced debitage ranged in counts from a single flake
(20CH433) to 14,900 flakes at the proto-historic component at the Summer Island site (20DE4).
The sites that produced debitage had a mean count of 816.3 flakes and a median count of 140.5
flakes. Only eight sites exceeded the mean count, and the standard deviation is 2468.2,
demonstrating the range and disparity in the assemblage sizes.
Raw material was documented at 41 of the sites (Appendix D). A recent study of the raw
material composition of debitage has demonstrated a relationship between raw material type and
the relative age of coastal archaeological sites in the Munising Bay area (Drake et al. 2009).
Based on this study, Archaic sites typically include 70 percent or more quartzite in their lithic
assemblage, whereas Woodland sites include less than 30 percent quartzite. The proportion of
quartzite on multicomponent sites (Archaic and Woodland) falls between 30 and 70 percent. In
fact, site 20AR338 which is explored in Drake et al.’s (2009) study illustrates the complexity of
extrapolating site age from the raw material composition of multicomponent sites. It is not clear
if this model holds true for interior sites or other geographic locales in the UP.
Chert comprised over 60 percent of the debitage assemblage on 32 of the 41 sites (78
percent of the sites where raw material was documented) and quartz was the best represented raw
material (over 40 percent of the debitage) on three sites, all of which are on Grand Island
(20AR348, 20AR353, and 20AR495). Quartzite was the most common (over 60 percent of the
debitage) raw material at four sites (9.8 percent of the sites). Two of these sites are on Grand

39

Island (20AR359 and 20AR400) and the other two (20DE459 and 20ST227) are located in the
Indian River drainage. Three of the four sites are known to be multicomponent (20AR359,
20AR400, and 20DE459). The results of this experiment suggest that Drake et al.’s (2009)
model may have application for sites outside the Munising Bay region
The reduction sequence of the debitage is recorded for only 15 LW sites, each of these
are on HNF lands (Appendix D). The first or primary stage of stone tool production is related to
the initial testing of raw material and the earliest stages of core preparation. The secondary or
middle stage of chipped stone tool production is typically associated with the manufacture of
usable flakes and tool blanks as well as the early stages of bifacial reduction. The final stage is
most closely associated with the later stages of core production, tool manufacture, tool
maintenance. Thus the reduction sequence can provide insight into the activities carried out on a
site (Robertson 1993). Sites with a balance of reduction stages were involved with all tool
preparation activities, whereas sites where one activity is overwhelming present may reflect that
stage of tool preparation and manufacture.
Two sites (20AR437 and 20MK261) have the highest percentage of late/final stage
reduction indicating that finishing and maintaining tools was an important activity, though not
the only activity, on these sites. Two sites (20AR398 and 20MK334) have the highest
percentage of the first stage of reduction suggesting that the initial stages of tool production was
an important activity on these sites. In fact, all the debitage from 20MK334 was from the initial
stage of testing and reduction. The second stage of production was the best represented type on
the remainder of the sites. Interestingly, 20MK261 and 20MK334 are situated proximate to one
another at the mouth of the Carp River in Mackinac County and possibly reflect two spatially
distinct activity areas of the same site.

40

3.1.3 Ground Stone Tools
One hundred and nineteen ground stone tools were collected at 24 LW sites or
components (Appendix D). The Juntunen site (20MK1) produced 31 ground stone tools, but
only seven could be securely placed in the three primary components (The Mackinaw, Bois
Blanc, and Juntunen Phases). The Scott Point site (20MK22) had 25 ground stone tools and
eight sites had a single ground stone tool.
The ground stone tool assemblages can be placed in nine functional categories or types as
well as an unclassified category. Hammer stones are the most common ground stone tool
appearing on 16 sites. Anvil stones and celts each appear on six sites and manos (grinding
stones) and net sinkers each appear on four sites. The remaining categories only appear on one
or two sites including: abraders (n=1); pestles (n=2); an adze (n=1); a piece of drilled slate (n=1),
and unclassified ground stone (n=2). The adze was found at the Juntunen site, but could not be
assigned to a specific component.
Hammer stones and anvils were likely used in chipped stone tool production, but could
also be used in food processing (cracking animal bone or nuts). Net sinkers were used in fishing.
Manos and pestles could be used in food processing as well as grinding other materials. The
celts and the adze were likely used for woodworking and/or chopping, but other functions are
possible.

3.2

Late Woodland Ceramics
Ceramics were recovered from 68 LW sites in the eastern UP (Appendix E). Sixty nine

components will be directly discussed here – two components at the Summer Island site (20DE4)

41

are presented in such a way that they can be presented as separate components, whereas elements
of some of the other sites, namely Juntunen (20MK1), Scott Point (20MK22), Cloudman
(20CH6), and 20MK169/457, are each multicomponent, but the details on the ceramics and
individual components is not clear enough to address the components separately in this instance
(Branstner 1995; Brose 1970; McHale-Milner 1998; McPherron 1967). The Juntunen site is
especially frustrating in that primary ceramic types (Mackinac, Bois Blanc, and Juntunen wares)
can be attributed to a component, but other varieties are not quantified in relation to specific
components (McPherron 1967). Seven of the sites have not produced any ceramics and another
six either include or likely include LW ceramics, but are either not well reported (or still being
reported on) or are sites with multiple components where the best studied component is not LW
(e.g., the Naomikong Point site [Janzen 1968]).
The 69 LW components include a minimum number of 2,308 ceramic vessels (MNV).
The overwhelming majority of these, 1,656 (71.9 percent), are from the Juntunen site (20MK1)
(McPherron 1967). The remaining 652 vessels were recovered from the remaining 68 LW
components. Thirty sites include a single vessel and two sites, aside from the Juntunen site,
include LW MNV counts of over 100 (n=136 at Cloudman [20CH6] and n=195 at Scott Point
[20MK22]).
As many as 29 broad typological classifications are represented with most sites including
one or two varieties and two sites, Juntunen (20MK1) and Getewaaking (20MK169/457), with
11 and 13 types respectively. The best represented ceramic types are Oneota-related wares
(present on 21 sites), Juntunen wares (n=17 sites), Mackinac wares (n=15 sites), Sand Point
wares (n= 8 sites), Bois Blanc wares (n=7 sites), and Iroquoian-related wares (n=6 sites).
Untypeable miniature vessels were recorded from six sites. For this discussion, an additional

42

category of “Wisconsinoid” ceramics was created subsuming early LW types best represented in
Wisconsin into a single category (Madison wares, Point Sauble wares, and Heins Creek wares).
The Wisconsinoid types appear on four sites.
There are some geographic trends associated with the ceramics (Figures 5 and 6). When
the ceramics are considered from a north/south perspective, there are 29 northern sites and 40
southern sites (42 percent and 58 percent respectively). Sand Point wares are more likely to
appear on northern sites, whereas Oneota-like, Bois Blanc, and Mackinac wares as well as
Miniature vessels are better represented in the south. The distribution of Juntunen and
Wisconsinoid ceramics are similar to the proportion of northern and southern sites. When
considered from an east/west perspective (32 eastern sites [46 percent] and 37 western sites [54
percent]), Sand Point and Wisconsinoid wares are better represented on western sites and the
other varieties are more easterly oriented. Bois Blanc and Mackinac wares are both strongly
associated with sites in the east.
100
90

Oneota

80

Juntunen

70

Bois Blanc

60

Mackinac

50

Sand Point

40

Miniature

30

Iroquoian

20

Wisconsinoid

10

Total Sites

0
North

South

Figure 5: Percentage of ceramics by type (North/South).

43

100
90

Oneota

80

Juntunen

70

Bois Blanc

60

Mackinac

50

Sand Point

40

Miniature

30

Iroquoian

20

Wisconsinoid

10

Total Sites

0
West

East

Figure 6: Percentage of ceramics by type (East/West).

Temporal trends are also apparent. The best represented ceramic types can be lumped
into early LW (~AD 600 to AD 1000) and late LW groups (~AD 1000 to AD 1600). For
example, Mackinac wares and Wisconsinoid wares are early LW types, whereas Oneota,
Juntunen, Bois Blanc, Sand Point, and Iroquoian-related wares are late LW (Brose 1970;
Dorothy 1980; Mason 1981; McHale-Milner 1998; McPherron 1967). Eighteen sites include
early LW ceramics and 37 components include late LW ceramics. Early LW sites are better
represented in the south and the east than the north or west (Figures 7 and 8). The spatial
distribution of late LW sites is proportionally similar to the distribution of LW sites in general.
This pattern suggests that early LW sites may have been focused on areas in the southeastern part
of the UP, such as the Straits of Mackinac, and that late LW sites were more evenly distributed
across the eastern UP.

44

80
70
60
50
Early
40

Late

30

Total Sites

20
10
0
North

South

Figure 7: Percentage of ceramics by age (North/South).
80
70
60
50
Early
40

Late

30

Total Sites

20
10
0
West

East

Figure 8: Percentage of ceramics by age (East/West).

3.3

Other Artifact Classes
In addition to the artifacts discussed above, a small number of sites included artifacts

made of different raw material types (bone and copper) as well as artifacts that were not used for
subsistence, tool maintenance, or other economic purposes such as pipes and beads (Appendix
F).
45

Pipes have been found at 11 LW sites/components (98 pipes and pipe fragments)
including all LW contexts at the Juntunen site. Pipes were made from clay (ten sites) and stone
(two sites) with one of the sites having both stone and clay pipes (20ST1). The Juntunen site
includes 65 pipes with ten associated with the Mackinac component, seven with the Bois Blanc
component, and nine with the Juntunen component. The remaining 39 could not be placed into a
specific component. The Getewaaking site (20MK169/457) produced two pipes. As mentioned
above, the Ekdahl-Goudreau site (20ST1) produced both clay and stone pipes, but counts are not
readily available. The remaining sites produced a single pipe or pipe fragment.
Copper artifacts were recovered from nine LW sites/components (141 artifacts) including
all LW contexts of the Juntunen site. Awls are the most common copper tool (n=66) and appear
on the most components (n=4). Copper beads and copper knives have been found at two sites.
The remaining sites have produced cones/projectile points (n=1), an effigy (n=1), a pin (n=1), a
ring (n=1), and a site with copper fragments.
Bone tools have been found at six LW sites/components (179 artifacts) including all LW
contexts at the Juntunen site. The Cloudman site (20CH6) and the Oneota/LW component of the
Summer Island site (20DE4) are the only other sites to produce bone artifacts. Like copper, bone
awls are the best represented tool type appearing on five LW sites (117 artifacts). Bone
harpoons have been found at four LW sites and projectile points at two sites. Needles, including
netting needles, have been found at two LW sites. Chisels and bone tubes have each been found
at one LW site.

46

3.4

Diversity Use Index
An underlying goal of this study is to determine if LW sites were used for different

purposes. One way to assess that question is through the material remains found at those sites.
As we have discussed, there is a range of materials found at LW sites in the eastern UP. The
scale and types of investigation at the sites, as well as the nature of the sites themselves, have led
to a disparate and diverse range of assemblages. The relatively high proportion (about a third) of
known LW sites which had only been explored through archaeological survey or limited test
excavation (10 m² or less [another third]) placed constraints on the interpretation of site function
(Appendix B).
The number and diversity of the artifact assemblages on the LW sites in the eastern UP
led to the adoption of a diversity use index (DUI; see Kvamme 1985). The index is based on the
assumption that different tools are used for different activities and that a greater diversity of tools
on a given site would reflect a greater range of activities. Conversely, a lack of tool diversity on
a given site could suggest a more limited range in activities. In a sense, the DUI is a simple
delineation addressing a greater or lesser range of activities on a site may help differentiate how
that site was used. This approach helped smooth the disparate data sets making it more
approachable, minimize bias, and facilitate interpretation.
Binford (1980) has conceptualized a framework to help understand hunter-gatherer
settlement and subsistence strategies as well as how one might think about archaeological site
formation resulting from these practices (see also Binford 1983). This model characterized
foragers as residentially mobile and collectors as logistically mobile, but Binford (1980:12)
states “… we are not talking about two polar types of settlement-subsistence systems, instead we
are discussing a graded series … .” Residential mobility refers to the movement of an entire

47

group from one location to another in pursuit of resources, whereas logistical mobility entails
smaller groups leaving and returning to a residential camp with resources (see also Holman and
Lovis 2008; Kelly 1992; Lovis et al. 2005; Whallon 2006). An expectation of such a system
would be that one might observe a greater diversity of activity at a residential camp and a lesser
degree of activities carried out at a logistical camp.
The DUI used in this study was calculated by multiplying the number of formal tools by
the number of morphological types. The tool categories included chipped stone tools, ground
stone tools, and ceramics (MNV). The resulting score for each site/component was used as a
scale to estimate the diversity of activities on each site (Appendix G). A high score reflects a
greater number and/or variety of tools, and a low score indicates the opposite. A greater number
and variety of tools is interpreted to represent a greater range and diversity of activities. The
DUI scores ranged from a score of one, for multiple sites, to a high of 3,024 for the Mackinac
Phase component of the Juntunen site (20MK1). The mean of the DUI spread is 153.3 and the
standard deviation 458.9.
An important caveat to this approach was discovered based on the different scales of
excavation on the individual sites. A correlation was noted between the amount of excavation
and the DUI score. Sites with higher scores typically had more excavation than sites with a low
DUI score (Correlation Coefficient [r] = 0.79). In other words, more excavation is directly
related to higher scores (e.g., Kintigh 1984). The Mackinac Phase component of the Juntunen
site, for example, included over 400 square meters of excavation and produced the highest score
(DUI=3,024). Conversely, many of the sites that produced a DUI score of one were known only
from archaeological survey and included less than one square meter of excavation (Appendices
B and G).

48

In an attempt to correct the DUI scores for the scale of excavation, the DUI score was
divided by the number of square meters excavated on a given site (DUIrev). In cases where the
site was known based on survey with a small number of shovel tests or limited surface
collection, a minimum of one square meter was used to calculate the scale of excavation. This
method effectively reduced the correlation (r = 0.25). The DUIrev scores range from one, for
multiple sites, to a high of 46.5 at 20MK90 (Appendix G). The mean of the DUIrev spread is 5.2
and the standard deviation 7.7. Site 20MK90 produced an original DUI score of 108, but only
included about 2.3 square meters of excavation. The DUI score of the Mackinac component on
the Juntunen site dropped from 3,024 to a DUIrev score of 6.9 because of the 441 square meters
of excavation.
The DUIrev score appears to address the relationship between more excavation and
higher DUI score. It creates a mean DUI and this is a useful measure for comparing sites. A
higher DUIrev might indicate a more intensive occupation with more activities (a higher range of
diversity per square meter of excavation), where as a lower DUIrev might indicate less activity.
However, it doesn’t address the relationship between the spatial extent, or size, of a site and the
potential DUI. It is possible that the same number of tasks and a comparable level of diversity
might be dispersed across a larger site area. In this instance, a larger area of excavation would be
required to recover the same range of activity. For example, the Juntunen site is much larger
(7432 m²) than site 20MK90 (500 m²). Less than one percent of site 20MK90 and only about 6
percent of the Juntunen site have been excavated (the mean amount of excavation for all the LW
sites in the eastern UP is 1.6 percent). The relatively low volume of excavation compared to the
spatial extent of each site raises the question as to whether the DUIrev score reflects the
density/intensity of occupation of the entire site or only that of the portion sampled.

49

The DUI and DUIrev scores were compared with the estimated spatial area of the sites
(m²) to explore this question (see Appendices B and G). Four sites that produced DUI scores did
not have estimated horizontal sizes (20DE17, 20MK3/11, 20MK53, and 20MK239). Each of
these sites had low DUI scores (≤ 5). The DUI scores had a weak correlation with site size (r =
0.37), and the DUIrev scores had virtually no relationship with site size (r = 0.04). This exercise
suggests that the horizontal extent of a site isn’t necessarily a factor in the assemblage diversity.
The two scales (DUI and DUIrev) were plotted against one another on a scatter plot
graph (Figure 9). Additionally, the mean of each set was calculated and three DUI categories
were established. The nine LW components (11.8 percent) that exceeded the mean in both scales
were characterized as having a high level of, or extended, diversity. Those components that
exceeded the mean in only one set or the other were classified as having intermediate diversity
(19 components [25 percent]). The remaining 48 LW components (63.2 percent) that did not
exceed the mean in either category were classified as having limited diversity.
The DUI categories presented above identify LW sites as having extended diversity,
intermediate diversity, or limited diversity in their assemblages. The sites with extended
diversity most likely represent residential sites. Residential sites would be expected to have a
longer duration of occupancy as well as larger populations that include mixed gender and age
groups that are involved with a wider number of activities (Binford 1980; 1983). Intermediate
diversity sites could also represent residential sites, albeit with a smaller population or shorter
occupation, or a logistical camp. Logistical camps are used for specialized purposes, by a
potential age and/or gender exclusive group, for a shorter period of time (Binford 1980; 1983). In
some instances, the same logistical camp location might be reoccupied and used by the same

50

64
32

DUIrev Score

16
Extended Diversity

8

Intermediate Diversity
4

Limited Diversity

2
1
1

10

100

1000

10000

DUI Score

Figure 9: Scatter Plot DUI and DUIrev scores.
group over multiple seasons or years for the same or different logistic activities. Finally, LW
sites with limited diversity likely represent logistical camps, although sites known from limited
excavation or as a result of archaeological survey may reflect limited diversity as a result of the
small sample size.
Not surprisingly, the nine sites with extended diversity have each been more extensively
studied and include: the three main components of the Juntunen site (20MK1) (McPherron
1967); the protohistoric component of the Summer Island site (20DE4) (Brose 1970); the
Cloudman site (20CH6) (Branstner 1995); the Scott Point site (20MK22); the Bark Dock site
(20CH95) (Dunham and Hambacher 2007); the Gete Odena site (20AR348) (Dunham and
Branstner 1995; Robinson et al. 1991; Skibo et al. 2004); and the Carp River site (20MK261)
(Dunham et al. 1993). There is little question that each of these sites can be characterized as
places where many activities were carried out by LW people.

51

The extended diversity sites are also geographically dispersed and all are in coastal
settings (Figure 10; see also Figure 9). The Juntunen and Carp River sites are in the Straits of
Mackinac area on Lake Huron. The Cloudman site is on Drummond Island near the head of
Lake Huron. The Summer Island site is located in Bay de Noc in Lake Michigan. The Scott
Point site is situated between Summer Island and the Straits of Mackinac on the Lake Michigan
shore. Gete Odena is on Grand Island in Lake Superior. Bark Dock is on Lake Superior on
Whitefish Bay.

Figure 10: The Locations of Late Woodland Sites with Extended Diversity.
The temporal affiliation of the extended diversity sites is such that Gete Odena, Juntunen,
Scott Point, and Cloudman each include both early and late LW components. The Carp River site
includes an early LW component, but no evidence of late LW activity. Finally, the Bark Dock
52

site and the Summer Island site include late LW components without evidence for early LW
occupations.
The intermediate diversity sites exceed the mean score of either the DUI or DUIrev, but
not both. Four sites exceed the mean for the DUI set and 15 exceed the mean for the DUIrev set.
The LW sites in the intermediate diversity category illustrate some of the shortcomings of the
two scales as well as how they complement one another. For example, the two sites with the
highest DUIrev scores, 20MK90 and 20DE296, are both coded as having intermediate diversity.
Both these sites have had very little excavation (2.3 and 3.4 m², respectively). Conversely, two
of the highest DUI scored sites in the intermediate category, the LW component of the
Naomikong Point site (20CH2) and the Getewaaking site (20MK169/457), each had a relatively
large amount of excavation (255 and 177 m², respectively) leading their DUIrev scores to be
reduced to one. The intermediate diversity category includes sites where many activities have
taken place (particularly those that exceeded the DUI mean) as well as sites that had more
intense occupations (particularly those that exceeded the DUIrev mean).
The intermediate diversity category includes 19 LW components, and five of these are
known only from archaeological survey (20CH171, 20DE7, 20DE93, 20DE333, and 20DE378).
Thirteen of the intermediate diversity sites are in coastal settings and six are in interior settings.
The coastal sites follow the same basic spatial distribution as the extended diversity sites with
four on Bay De Noc, four in the Mackinac Straits area, three on Grand Island or Munising Bay,
one on Whitefish Bay, and one between the Straits and Bay de Noc on Lake Michigan. The
interior sites are situated in locales which drain towards Bay de Noc (n=3), northern Lake
Michigan and/or Bay de Noc (n=2), and the Straits of Mackinac (n=1).

53

Six intermediate diversity sites include evidence for early and late LW activity. Two sites
only include evidence for early LW occupation (20DE7 and 20MK90). Four sites only have
evidence for late LW components (20AR437, 20DE296, 20DE333, and 20ST109/110). The
remaining sites did not produce evidence that would allow anything more than a LW assignation.
The limited diversity category includes 48 LW components or 63.2 percent of the LW
sites. This category represents sites with the fewest activities represented. Based on the DUI
scores, 16 components (21 percent) scored one (a single tool in a single category). Nine of these
are known from archaeological survey and seven from small scale excavation. When the DUIrev
scores are examined, there are 27 sites (35.5 percent) that scored one (including all 16 of the sites
that scored a DUI of one). Eleven of these are known from archaeological survey and the other
16 from small scale excavations.
The sites in the limited diversity category are well distributed across the eastern UP. A
much higher proportion of sites are located in the interior (18 of the 48 sites [37.5 percent]) than
in the preceding categories. Twenty eight sites produced evidence for site age with 6 including
evidence for both early and late LW activity, three with evidence for only early LW use, and 19
with only evidence for late LW occupation. One interior site was used in both the early and late
LW, and seven sites in the interior only had evidence for late LW activity.

3.5

Subsistence Remains
In addition to the artifact classes discussed in the previous sections, some of the LW sites

produced subsistence remains. Subsistence remains represent the plants and animals that were
eaten or used by the inhabitants of these sites. These remains have the potential to inform about
LW diet and the ecosystems LW people were obtaining these resources from as well as the time
of year they were using them. The purpose of this section is to provide a summary of the
54

subsistence remains recovered in the eastern UP to provide data that may enhance the DUI
analysis presented above as well as the environmental data presented in the subsequent chapters.
As noted in Chapter 2.2, there is not a proportionate data set for subsistence remains in
relation to all LW assemblages. The two primary constraints to subsistence remains reflect
excavation strategy and taphonomic factors (poor preservation due to acidic soils, severe freeze
thaw cycles, and weathering [T. Martin et al. 1993]). In regard to excavation strategy, the
primary limitations reflect the lack of consistency in how the remains were collected and, if they
were collected at all. Subsistence remains are not typically recovered from archaeological
surveys. They are more typically recovered from archaeological excavations, but provision to
collect such data is not always part of the excavation strategy. In this instance, earlier
excavations such as at Juntunen (20MK1) and Summer Island (20DE4) recovered subsistence
remains as part of the standard excavation (Brose 1970; McPherron 1967). This typically
involved screening soil with ¼ in (0.64 cm) screen, “…except when particular caution was
indicated …” (McPherron 1967:25). When one considers that most seeds and fish bone are
smaller than a quarter inch, recovering a reasonable sample of subsistence remains was difficult.
In more recent excavations, finer scale sampling techniques such a flotation were used
specifically to recover such remains (see Anderton et al. 1991; Branstner 1995; Dunham and
Branstner 1995; Dunham et al. 1993). The finer recovery techniques mitigated some of the bias
of the earlier excavations, but only when those techniques were used. More recently, residues
adhering to pottery and FCR have been recovered (see Kooiman 2012; Skibo et al. 2009). This
approach has great potential. Not only does it potentially better address recovery techniques, it
has the potential to mitigate some of the problems with poor preservation of bone and plant
remains (Malainey 2007).

55

3.5.1 Faunal Assemblages
Faunal remains have been recovered from 43 LW sites, or individual components of LW
sites, in the eastern UP. Twenty seven of these have produced identifiable faunal remains
(Appendix H). These include mammals, fish, birds, reptiles, and a small number of mollusks and
gastropods. The numbers of identified species varies significantly by site and typically reflect
the scale of excavation at the site, although taphonomic factors and recovery techniques are also
critical factors. This discussion focuses on the number of identified species, as opposed to other
metrics.
Two LW sites (30AR338 and 20CH2) in the study area included lipid analyses which
provided basic information on faunal species cooked in ceramic vessels. One of the LW vessels
from the Naomikong Point site (20CH2) produced a high amount of animal fat and it is unlikely
that any of the Naomikong vessels were used to cook fish (Kooiman 2012). Some animal fat
was noted at 20AR338 associated with a likely LW vessel, however the lipid remains from that
site were mostly plant based (Skibo et al. 2009). For the purposes of this discussion, the focus
will be on the identified faunal data.
The LW components with identifiable fauna range from one identified species to 34
identified species from four taxonomic classes (mammal, fish, reptile, and bird) (Figure 11). The
Mackinac Phase component of the Juntunen site (20MK1) includes the highest number of
species and includes: 11 species of mammal; 15 species of fish; seven species of bird, and one
species of reptile. Each of the three LW components at the Juntunen site, a mixed LW
component from 20MK169/457 on Mackinac Island, the Bois Blanc component of the Scott
Point site (20MK22), and 20DE296 on Big Bay de Noc had the highest number of identified

56

20ST109/110
20ST1
20MK90
20MK61
20MK54
20MK261
20MK24
20MK22 (Mackinac)
20MK22 (Juntunen)
20MK22 (Bois Blanc)
20MK169/457 (Mixed LW)

20MK169/457 (Early)
Mammal Diversity

20MK169/457 (Bois Blanc)
20MK1 (Mackinac)

Fish Diversity

20MK1 (Juntunen)

Bird Diversity
Reptile Diversity

20MK1 (Bois Blanc)
20DE75
20DE4 (pHST)
20DE4 (O/LW)
20DE296
20DE188
20CH95
20CH6
20CH238
20AR437
20AR359
20AR348
0

5

10

15

20

Figure 11: LW components with identifiable fauna.

57

25

30

35

20MK1 (Mackinac)
20MK1 (Bois Blanc)
20MK169/457 (Mixed LW)
20MK1 (Juntunen)
20MK22 (Bois Blanc)
20DE296
20MK22 (Mackinac)
20MK22 (Juntunen)
20CH6
20AR348
20MK261
20MK169/457 (Early)
20MK169/457 (Bois Blanc)
20ST1
20ST109/110
20MK54
20MK61
20DE4 (pHST)
20AR437
20DE4 (O/LW)
20AR359
20DE188
20CH95
20MK90
20MK24
20CH238
20DE75
0

20

40

60

80

100

120

140

Faunal Diversity Score

Figure 12: Faunal Diversity Score.
species. A diversity index for faunal species was created by multiplying the number of identified
species by the number of varieties (mammal, fish, bird, and reptile). The Mackinac Phase
component of the Juntunen site has the highest diversity score of 136 (Figure 12; Appendix H).
Using the number of identified species and diversity score, some general observations
can be made about the LW sites in the eastern UP (Table 1). When all the components are
considered, the mean number of identified mammals and fish is higher than the mean number of
identified reptiles or birds. An equal mean number of fish and mammals are represented and
58

birds are nearly twice as well represented as reptiles. It can also be observed that a wider range
of species are present on coastal sites as opposed to those in the interior (Coastal Mean Diversity
50.0 and Interior Mean Diversity 8.8), although the general relationship between the mean
number of represented species remains consistent (fish and mammal are about the same, and the
mean of birds is higher than reptiles). The overall discrepancy probably reflects the fact that
there are 22 coastal sites and only five interior sites with identifiable fauna. Likewise, nearly all
the coastal sites with identified fauna are multicomponent, multiple occupations and all the
interior sites appear to be single component and probably much shorter duration occupations. In
other words, bigger sites with more activity result in a wider range of represented species. The
consistency in the general proportions of represented species suggest a similarity in use patterns
of faunal species and the differences are more likely a result of the scale of use of the sites
themselves.
Mean ID
Species

Mean ID
Mammal

Mean ID
Bird

Mean ID
Reptile

Mean ID
Fish

Mean
Diversity

All LW Sites

12.0

4.7

1.7

0.9

4.7

42.4

Coastal

13.9

5.4

2.0

1.0

5.4

50.0

Interior

3.8

1.8

0

0.4

1.6

8.8

Early

13.0

4.9

1.8

0.9

5.3

45.9

Late

12.2

5.1

1.6

1.0

4.6

42.9

East

14

5.2

2.2

1.2

5.4

51.7

West

8.0

3.7

0.7

0.4

3.2

23.6

Fauna

Table 1: Mean number of identified fauna and mean diversity score.

When sites with early LW components are compared to sites with late LW components
there are similarities and differences. The early and late categories overlap one another with
twelve sites having early LW components and 20 having late LW components (six of these have

59

both late LW and early LW components). The mean number of identified species is similar for
early LW and late LW (n=13.0 and 12.2, respectively) and the early LW appears to have a
slightly greater mean diversity than the late LW (45.9 early LW and 42.9 late LW). The most
striking difference is a modest shift in the ratio of mammals to fish. For all components, this is a
1:1 relationship, whereas it is a ratio of 4.9:5.3 for the early Late Woodland and a ratio of 5.1:4.6
in the late Late Woodland. While not a statistically significant difference (χ² = 0.07, df = 1, p =
0.449), it offers the potential that late LW peoples were using a wider variety of mammals than
early LW peoples, and that early LW peoples made use of a wider variety of fish than late LW
people. A pattern of greater use of mammals has been observed at the late LW components at
the Scott Point and Juntunen sites (T. Martin 1982; Cleland 1966).
When the LW components are separated by the broad geographic categories of east and
west a parallel set of trends can be observed. There are 18 components with identifiable faunal
remains in the eastern half of the study area and nine in the western half. The mean number of
species as well as the mean diversity score is significantly higher for the eastern sites than the
western sites (14/7.8 and 51.7/23.6, respectively). This likely reflects the coastal/interior
distribution of the components as four of the nine western sites are in the interior and only one of
the 18 eastern sites is in the interior. However, where the general mean proportion of mammals
to fish was about 1:1 for both coastal and interior sites, western sites have a slightly higher mean
number of mammal species than fish species (3.7:3.2), and eastern sites have a slightly higher
proportion of fish compared to mammal species (5.2:5.4). This may indicate regional difference
in mammal to fish use, but it also may reflect a higher proportion of early LW components in the
eastern portion of the study area (11 of the 18 sites with identifiable fauna in the east have an
early LW component and only one of the nine in the western portion has early LW components).

60

In addition to examining the diversity of species represented at individual components
and sites, the relative ubiquity of individual species were examined across the twenty seven LW
components that produced identifiable fauna. This discussion centers around how many
components include a given species, in a presence or absence sense. This is not an attempt to
determine the relative importance of individual species to the diet, but rather to determine what
resources were appearing on the most LW sites or LW components. This, in turn, may provide
an insight into the habitats hunted and fished to procure these animals.
Beaver appears on the most sites (n=22) followed closely by lake sturgeon which appears
on 20 sites. Figure 13 is a bar graph that illustrates those species that appear on at least eight (ca.
30 percent) of the 27 LW components. These 15 species are considered the most ubiquitous.

Other Notable Species

Best Represented Species

Figure 13 also includes seven species that are notable and will be discussed later in this section.
Beaver
Lake Sturgeon
Walleye
White-tailed deer
Lake trout
Wolf/dog
Sucker
Moose
Northern pike
Bass (smallmouth)
Lake whitefish
Common loon
Snapping turtle
Painted turtle
Black bear
Whitefish family
Caribou
Passenger Pigeon
Bald eagle
Ducks, sp.
Canada goose
Wild turkey
0

5

10

Figure 13: Best represented and other notable species.
61

15

20

25

The best represented species are primarily mammals and fish along with one species of bird
(loon) and two varieties of turtle (reptiles). Whitetail deer, along with sturgeon and beaver, are
well represented. Fall spawning lake trout and whitefish are also present on the list with lake
trout recovered from 13 components and whitefish from 11. Spring spawning walleye, sucker
and pike were each found on more than 11 components. Moose, the largest animal in the region,
was recovered from 12 components.
Whitetail deer, sturgeon, and beaver are common on LW sites in the region and expected
because of the size or density of the bone (deer and beaver) or the diagnostic character of
sturgeon dermal plates (or scutes). These also happen to be the most ubiquitous species on
eastern UP LW sites.
The differences in diversity of species between coastal and interior sites are also apparent
in reviewing the most ubiquitous species. Figure 14 shows the relative percentage of each of the
species by coastal and interior setting. Each species, with the exception of northern pike, appears
on a higher proportion of coastal sites than interior sites. Further, bear, loon, whitefish, bass,
wolf/dog, lake trout, and walleye only appear on coastal sites in the eastern UP. While these
differences may reflect different subsistence patterns on the interior, it more likely reflects the
smaller number of interior sites as well as the more intensive use of coastal sites over longer
periods of time.
Lake trout and whitefish are endemic to the Great Lakes, but not the UP’s interior lakes.
For these fish species to appear in interior assemblages would require people to bring them to the
interior locales. The absence of walleye, smallmouth bass, bear, loon, and wolf/dog is curious,
as there is no reason these species should not be present on inland sites. In fact, B. A. Smith
(1996) identified bear as one of the more important species for people on Lake Superior Interior
62

100
90
80
70
60
50
Interior

40

Coastal
30
20
10
0

Figure 14: Relative percentage of the best represented species by coastal & interior setting

sites in Ontario, an area adjoining the eastern UP to the northeast. While loons would have been
present on inland lakes, it is possible that loons were more likely captured when they entangled
themselves in nets (Cleland 1966; McPherron 1967; B. A. Smith 1996).
Figure 15 shows the relative percentage of the most ubiquitous species comparing early
LW and late LW components. As the graph indicates, late LW sites include a slightly higher
proportion of most species, although early LW components have a higher proportion of lake
trout, wolf/dog, walleye, sturgeon, and beaver. When the specific differences are examined,
wolf/dog, walleye, and sturgeon appear on a higher proportion (> 9 percent) of components in

63

90
80
70
60
50
40
Late LW
30

Early LW

20
10
0

Figure 15: Relative percentage of the most ubiquitous species comparing early LW and
late LW components.
the early LW, and bear, bass, pike, and deer appear in a higher proportion (> 9 percent) in the
late LW. The increased percentage of late LW components that include deer, bear, and moose
(6.5 percent) may reflect a greater reliance on mammals for subsistence, a trend also suggested
by the diversity discussion above.
A comparison of the relative percentage of the most ubiquitous species by eastern and
western portions of the project area also reveals some noteworthy differences (Figure 16). First,
caribou and wolf/dog only appear on eastern sites. Second, all the species that are proportionally
better represented in the western part of the project area (deer, bear, pike, and painted turtle) are
less than 6 percent better represented than on the eastern sites. Third, lake trout and whitefish
are nearly 40 percent better represented on eastern sites. Fourth, moose,

64

90
80
70
60
50
40
East
30

West

20
10
0

Figure 16: Relative percentage of the most ubiquitous species by east and west.
walleye, snapping turtle, and loon are each over 10 percent better represented on eastern
components.
As noted above, some of these differences may relate to the higher proportion of larger,
multicomponent coastal sites in the eastern region leading those sites to have a wider diversity of
species represented. The larger, multicomponent sites are more likely to have longer duration
and more repeated occupations (seasonal and inter-annual), have larger populations of mixed age
and gender allowing for a larger range and greater diversity of resources collected. Additionally,
there may be factors relating to habitat and animal range (caribou), variation in subsistence
strategy such as a greater emphasis on fall spawning lake trout and whitefish in the east (this may
also account for loon), or different cultural practices such as ritual use of dogs (see below).
In addition to showing the most ubiquitous species, Figure 13 also includes a seven other
notable species recovered from LW components in the eastern UP. These include fish remains
65

that were identified as relating to the whitefish family (Coregoninae), but that could not be
specifically identified to species. While these could represent shallow water spawning cisco or
another whitefish relative, they could also represent lake whitefish which could increase the
number of sites where that species appears by as many as two. Caribou is noted because it
appears on seven sites (25.9 percent of the sites) and is a large mammal. Canada goose and
ducks are noted because they appear on so few LW sites in the eastern UP despite their ubiquity
today. The identification of wild turkey at Juntunen (MNI=6, five from the Bois Blanc
component) is noteworthy because it is probably outside its native range suggesting they were
brought to the site (Cleland 1966; Brewer et al. 1991).
Bald eagle is recorded here because it likely does not represent a food resource, but rather
is related to cultural practices. Eagle burials were recorded as part of the Juntunen component of
the Juntunen site and an eagle talon was recovered from 20DE296 (Cleland 1996; T. Martin
1991). A dog burial and dog remains in an apparent medicine bundle were recovered from the
Juntunen site as well, suggesting an ceremonial or ideational role for dogs, although butchered
dog remains, for apparent subsistence, were also recovered from the Juntunen site, site
20MK457, and the Cloudman site (20CH6) indicating their role as a food resource (Cleland
1966; T. Martin and Perri 2011; Cooper 1996).
The shells of turtle species, such as the painted turtle, appear to have been used culturally
to make rattles or containers (Cleland 1966; T. Martin 1980; 1991; T. Martin and Perri 2011).
This doesn’t mean that turtles weren’t used for food as well, only that they were also used for
non-subsistence cultural purposes.
B. A. Smith (2004) has posited that the increased reliance on fall spawning whitefish and
lake trout began around AD 800 at the Juntunen site, after AD 1100 in northern Lake Michigan

66

basin, and as late as AD 1400 in the rest of the Upper Great Lakes region. Of the 22 LW
components on coastal sites in this study that included faunal remains, fifteen (68 percent)
included fall spawning fish (Table 2). Five components with fall spawning fish remains could
not be attributed to an early or late LW context, or included evidence for each of these periods.
Three are early LW in age (pre AD 1000) and seven are late LW (post AD 1000). If this data
were presented as percentages, then 50 percent of the early LW coastal components include fall
spawning fish and 70 percent of the late LW sites do as well. Despite the small size of the
eastern UP sample, this generally supports B. A. Smith’s (2004) observation that the fall fishery
increased in importance as a subsistence resource over the course of the LW period.

Site

Early LW

Late LW

Lake Trout

Whitefish

20AR348

E

L

x

-

20AR359

-

L

-

x

20DE296

-

L

x

-

20MK1 (Mackinac)

E

-

x

x

20MK1 (Bois Blanc)

-

L

x

x

20MK1 (Juntunen)

-

L

x

x

20MK22 (Mackinac)

E

-

x

-

20MK22 (Bois Blanc)

-

L

x

x

20MK22 (Juntunen)

-

L

x

x

20MK54

E

L

-

x

20MK61

E

L

x

-

20MK169/457 (Early)

E

-

x

x

20MK169/457 (Bois Blanc)

-

L

x

x

20MK169/457 (Mixed LW)

E

L

x

x

20ST1

E

L

x

-

Table 2: Fall Spawning Fish Recovered from Coastal Sites.

67

4

3

2

1

0
Michigan

Straits
Early

Superior

Michigan

Straits

Superior

Michigan

Late

Straits

Superior

Mixed

Figure 17: Number of sites with fall spawning fish by broad region.
The evidence is less clear from a geographic perspective. The sites with fall spawning fish
remains are located in three broad geographic areas: northern Lake Michigan (including Bay de
Noc); the Straits of Mackinac; and Lake Superior (Figure 17). Two early LW components in the
Straits area and one on northern Lake Michigan included fall spawning fish remains. Three sites
on the Straits, three on northern Lake Michigan, and one on Lake Superior included fall
spawning fish remains in the late LW. Of the five sites that could not be placed in either the
early or late LW, one is located on northern Lake Michigan, three on the Straits, and one on Lake
Superior. Thus, evidence of the fall fishery is best represented in the Straits region in all periods,
and the use of fall spawning fish is seemingly more prevalent in each region in the late LW.
When the discussion is expanded to include the most ubiquitous spring spawning fish
species, sturgeon, walleye, and sucker, the data demonstrates that spring spawning fish remains
appear on more coastal sites than fall spawning fish (19 of 22 sites with faunal remains [86
percent]). Five early LW sites (83 percent) have spring spawning fish and eight late LW sites
(80 percent) include spring spawning fish (Figure 18). The remaining six sites, with unidentified
68

9
8
7
6
5

Fall Spawning

4

Spring Spawning

3
2
1
0
Early (6 sites)

Late (10 sites)

E/L or Unk. (6 sites)

Interior (5 sites)

Figure 18: Number of sites with spring and fall spawning fish
or mixed LW components, also include spring spawning fish remains. Thirteen sites included
both fall and spring spawning fish which is 86.7 percent of the 15 sites that include fall spawning
remains. Spring spawning fish remains are also present on interior sites as well (3 of 5 interior
sites with faunal remains) which further illustrates their greater ubiquity in LW subsistence
practices.
The relative importance of large herbivores (whitetail deer, moose, and woodland
caribou) can also be explored in comparison with fall and spring spawning fish (Figure 19).
Large herbivores appear in 16 of the 22 (72.7 percent) coastal faunal assemblages (one more than
fall spawning fish) and three of the five interior assemblages. As such, large game is more
ubiquitous on LW sites than fall spawning fish and less ubiquitous than spring spawning fish.
Ninety percent of the late LW sites include large game in their faunal assemblages, as compared
to 80 percent spring spawning fish and 70 percent fall spawning fish (Figure 20). Only 50
percent of the early LW sites include large game which supports the observation discussed above
that mammals are better represented on late LW sites than early LW sites (Figure 3.5-20).

69

The most ubiquitous species appearing on LW sites in the eastern UP have a fairly
limited range of preferred habitats. All the fish species as well as the turtles, beaver, moose, and
loon each live in the water or lands adjacent to lakes, rivers, streams, or marshes (Tables 3
20
18
16
14
12

Fall Spawning

10

Spring Spawning

8

Big Game

6
4
2
0
Interior

Coastal

Figure 19: Comparison of fall spawning fish, spring spawning fish, & big game by count.
10
9
8
7
6

Fall Spawning

5

Spring Spawning

4

Big Game

3
2
1
0
Early (6 sites)

Late (10 sites)

E/L or Unk. (6 sites)

Interior (5 sites)

Figure 20: Comparison of fall spawning fish, spring spawning fish, & big game by age.
and 4). The remaining species – deer, bear, and the canines need access to water for drinking,
but are not constrained to the water itself or littoral zones. Most of the species are also available
70

throughout the year. Further, beaver, moose, and deer require habitats with early successional
growth as part of their range. Thus, the majority of the most ubiquitous species habitat
preferences would place them in locations that are most likely to include LW sites (within 240 m
of a source of water and/or in mixed pine habitats [see Chapter 4]). This would also pertain to
dogs whose primary habitat is in association with human settlement.
Despite these commonalities, there are some differences. The first relates to seasonal
availability. While most of the species are available throughout the year, there are times when a
given species is “more available,” or available in greater numbers than in others. This is well
illustrated by spring spawning fish such as sturgeon, walleye, sucker, and northern pike. While
each of these could be conceivably caught at anytime during the year, they all appear in larger
numbers and greater density during their spring spawning runs (Becker 1983; Cleland 1982; B.
A. Smith 1996). Spawning takes place in shallow waters where the fish can be speared or
caught in dip nets. Fish running in streams can be caught in weirs or traps as well as in nets. A
local legend relates that a person could walk across the Sturgeon River near Nahma (Bay de
Noc) across the backs of sturgeon during the spawning run (Dodge 1973:91).

71

Species

Preferred Habitat

Seasonal Availability

Reference

Sturgeon

Great Lakes, inland lakes,
rivers/streams

Year round, although
concentrated during spring
spawning

Becker 1983; Cleland
1982

Walleye

Great Lakes, inland lakes,
rivers/streams, and
marshes adjoining streams
and lakes

Year round, although
concentrated during spring
spawning

Becker 1983

Lake Trout

Great Lakes

Fall spawning

Becker 1983; Cleland
1982

Lake Whitefish

Great Lakes

Fall spawning

Becker 1983; Cleland
1982

White Sucker

Great Lakes, inland lakes,
rivers/streams

Year round, although
concentrated during spring
spawning

Becker 1983

Northern Pike

Great Lakes, inland lakes,
rivers/streams, and
marshes adjoining streams
and lakes

Year round, although
concentrated during spring
spawning

Becker 1983

Table 3: Habitat and seasonal summary for the most ubiquitous fish.

72

Species

Preferred Habitat

Seasonal Availability

Reference

Beaver

Early succession growth
adjacent to slow moving
streams, rivers, and inland
lakes

Year round, although
ethnographic sources
indicate late winter &
early spring as optimal
time

Baker 1983; B. A. Smith
1996

Whitetail Deer

Warm season: mosaic of
forest edges, early
successional growth, and
forest openings. Cold
season: Lowland conifer
swamp (esp. white cedar)

Year round, although they
yard in larger groups in the
winter

Baker 1983; Van Deelan et
al. 1996

Moose

Subclimax forest and
herbaceous openings
adjacent to swamps,
marshes, lakes

Year round

Baker 1983

Woodland Caribou

Mature coniferous forest
adjacent to swamps and
bogs

Year round

Baker 1983

Black Bear

Multiple environments

Year round, although
black bears hibernate from
late November-early April.
Ethnographic sources
indicate bears were hunted
while in hibernation

Baker 1983; B. A. Smith
1996

Wolf

Multiple environments

Year round

Baker 1983

Dog

With humans

Year round, although
enthnographic sources
note ritual use in midwinter

B. A. Smith 1996

Table 4: Habitat and seasonal summary for the most ubiquitous mammals.

73

3.5.2 Floral Remains
Macrobotanical floral remains have been recovered from fifteen components from
fourteen LW sites in the eastern UP (Appendix I). The floral remains include carbonized seeds,
nutshell, nutmeats, wood charcoal, as well as other plant remains such as aquatic tuber and birch
bark. At least 30 types of nuts and seeds are represented and a dozen varieties of tree. The
approach taken here is to consider ubiquity, or in how many components is a given species
represented. Few of the sites have produced large nut and seed assemblages, the Cloudman site
(20CH6) and the Juntunen site (20MK1) are exceptions, so this seems a useful strategy to
ascertain broader evidence of plant use in the eastern UP.
Lipid residues extracted from the fabric of pottery has also indicated plant use at a site on
Grand Island (20AR338) and at the LW component of the Naomikong Point site (20CH2)
(Kooiman 2012; Skibo et al. 2009). The results of these analyses are interesting, but do not
provide a lot of insight into specific plant remains. A LW vessel from the Naomikong Point site
had been used to cook animals and low fat content plants (Kooiman 2012:163). The apparent
LW vessel examined from 20AR338 produced evidence for nut oil, likely acorn (Skibo et al.
2009).
Twelve sites, including 13 LW components, produced carbonized seeds, nutshell and/or
nut meats. Eight sites produced wood charcoal. The LW components with floral remains range
from one identified species to 17 identified species (Figure 21). The two sites with the highest
number of identified plant remains have no identified wood charcoal (20CH6 and 20MK1).
Wood charcoal identification was not carried out for all sites including 20CH6 and 20MK1.

74

18
16
14
12
10
8

Diversity Seed/Nut Remains

6

Diversity Charcoal Remains

4
2
0

Figure 21: Diversity of floral assemblages (seed/nut and charcoal) by site
Acorn is the best represented nut or seed appearing on 46 percent (6 of 13) of the LW
components that produced these remains. Hazelnut and cherry appeared on 38.5 percent of the
sites (5 of 13). All the other seed and nut remains appear on less than a quarter of the sites.
Notable among these were plants that were important food resources to the south (southeast and
southwest) of the eastern UP including: maize which appeared on three sites (23 percent); wild
rice on one site; squash on one site; and likely butternut (Juglans sp.) on one site. In addition to
these resources, two sites also produced evidence for aquatic tubers although the species could
not be identified.
Acorns, hazelnut, and butternut are mast resources and nuts have played an important
role is human subsistence in eastern North America throughout prehistory (Gardner 1997; Scarry
2003; Yarnell 1964). Acorns and hazelnuts, along with beech nuts, are the best represented nuts
in the eastern UP, whereas butternut is more localized and rare (Comer et al. 1995; Dunham
2009; Voss and Reznicek 2012). Hazelnuts and butternut are oily nuts and acorns are starchy
(Kuhnlien and Turner 1991; Scarry 2003). This difference has nutritional outcomes with acorns
75

comparing more favorably to grains, such as maize or wild rice, than to other nuts (Dunham
2009; Kuhnlien and Turner 1991; Scarry 2003). Butternut (Juglans cinerea) is currently the only
variety of Juglans present in the eastern UP and is recorded in Chippewa, Delta, and Mackinac
County as well as appearing in pre-European settlement forest descriptions (Comer et al. 1995;
McPherron 1967; Voss and Reznicek 2012). Its presence on the Cloudman site, 20CH6, with
one of the most diverse botanical assemblages, is not a surprise (Branstner 1995; Egan-Bruhy
2007).
The squash and maize represent cultigens which were important crops to the south of the
eastern UP (Hart and Lovis 2012; McElrath et al. 2000; Parker 1996). Prehistoric maize is
assumed to require 140 frost free days to produce a reliable subsistence crop (Demeritt 1991;
Hart and Lovis 2013; Yarnell 1964). As noted in Chapter 2.1, areas in the Lake Michigan and
Lake Huron littoral zone currently exceed 140 frost free days (Eichenlaub et al. 1990; O’Shea
2003). All the sites where maize was recovered as well as Summer Island (20DE4) where the
squash was found currently exceed 140 frost free days.
Wild rice was an important resource in the western Upper Great Lakes and parts of
Ontario (Jenks 1900; Johnson 1969; Mather and Thompson 2000; Vennum 1988). It only
appears on a single site in the eastern UP, the Cloudman site (20CH6), and only as a single
macrofossil (Egan-Bruhy 2007). The Cloudman site is situated in close proximity to a modern
wild rice patch and generally reliable nineteenth century historical records place wild rice in the
general vicinity as well (Dunham 2008). The recovery of a single grain is not conclusive
evidence that the inhabitants of the Cloudman site collected the wild rice locally, it may have
come to the site through exchange, but it raises the possibility that wild rice was used in the
eastern UP.

76

Also notable was the low incidence of fleshy fruits with the exception of cherry.
Blueberries, raspberries, and grapes as well as wild plums only appeared on one or two of the
sites that produced edible plant remains. Even in the case of cherries, different varieties (pin,
choke and unspecified) were lumped together which achieved a higher number of total sites.
One would expect to see a higher proportion of fleshy fruits in these assemblages based on their
prominence in the ethnographic literature (Densmore 1974; Dunham 2000a; Kuhnlien and
Turner 1991; Moncton 1992). One factor that may explain this is the processing of fleshy fruits,
specifically drying, did not involve heating with fire, thus limiting the potential for carbonization
of the seeds (Dunham 2000a).
Site 20MK1, the Juntunen site, includes three distinct components as well as transitional
zones and the majority of the floral remains were found in occupation strata assigned to the
Juntunen Phase component (McPherron 1967; Yarnell 1964). Only maize, hazelnut, birch bark,
and an unidentified tuber were identified in earlier contexts. Eleven of the 15 maize kernels
(73.3 percent) that were found in assigned components were found in the Juntunen component.
The Cloudman site, 20CH6, is a multicomponent site with Middle Woodland, Late
Woodland, and early historic components (Branstner 1995). The floral assemblage was fairly
extensive, but the remains discussed herein are limited to LW features. Five varieties of fleshy
fruit, 3 varieties of nuts, wild rice, and maize were all recovered (Appendix I; Egan-Bruhy 2007).
The charcoal remains provide insight into the types of wood burned in fires by LW
people as well as species that occurred in closer proximity to occupations (Appendix I). The best
represented type of wood was birch (Betula spp.) which is present on 62.5 of the sites with wood
charcoal. Maple (Acer spp. and A. saccarum) and ash (Fraxinus spp. and F. americanus) are

77

present on half of the sites. Pines (Pinus spp., P. strobus, and P. resinosa) appear on three sites.
None of the other varieties of wood charcoal are present on more than one or two sites.
A couple of unusual wood charcoal specimens are worth noting. Two sites (20DE296
and 20MK54) produced hickory (Carya spp.) which is an uncommon species in the eastern UP
(Voss and Reznicek 2012). Hickory is an important mast resource to the south and raises the
potential for use of this mast source in the region despite hickory not appearing in the nut
remains above (Gardner 1997; Scarry 2003). Shagbark hickory (C. ovate) is recorded in the
recent past in Delta and Mackinac counties which correspond to the locations of these two sites
(Voss and Reznicek 2012). The Delta County shagbark hickory is considered a natural
occurrence, but no additional information is available for the specimen from Mackinac County
(Voss and Reznicek 2012:650). The other unusual specimen is identified as swamp white oak
(Quercus bicolar) at site 20MK54 (Brunette cited in Fitting and Clarke 1974). Swamp white oak
is not thought to extend north of mid-Michigan raising the possibility that the specimen was
misidentified (Voss and Reznicek 2012).

3.6

Discussion
One of the goals of this study is to determine if LW sites were used for different

purposes. The DUI scores provide a useful means for exploring this question. In this chapter,
three DUI categories were created: extended diversity, intermediate diversity, and limited
diversity (Appendix G). The sites classified as extended diversity most likely represent
residential sites. Intermediate diversity sites may also represent residential sites, albeit with a
smaller population or shorter occupation, or a logistical camp. Finally, LW sites with limited
activity scores likely represent logistical camps.

78

Most of the sites examined were scored as having limited diversity (n=48 [63.2 percent])
and only 9 components (11.8 percent) were determined to be extended diversity. The nine
components with extended diversity can each be characterized as places where many activities
were carried out by LW people. The extended diversity sites are geographically dispersed across
the eastern UP and each are situated along a Great Lakes Shoreline (see Figure 10). Two of the
components are early LW, 4 are late LW, and three of the sites include mixed deposits. The two
early LW components are in the Straits of Mackinac region and the 4 late LW components are
located on Lakes Michigan, Huron, and Superior. The extended diversity sites are also more
likely to include floral and faunal remains than the other sites and two-thirds of them also include
subsurface features (Figure 22). These sites are likely residential locales with longer duration of
occupation as well as larger populations that include mixed gender and age groups that are
involved with a wider number of activities. The extended diversity sites likely represent the
seasonal aggregation sites described by Cleland (1982) where spring and/or fall fishing took
place. Each of the extended diversity sites, with the exception of the proto-historic component of
20DE4, includes spring or fall spawning fish remains (Appendix H).
Intermediate diversity sites could also represent residential sites, albeit with a smaller
population or shorter occupation, or logistical camps. As residential sites, they may represent the
locales occupied during periods of population dispersal, such as cold season camps, as described
by Cleland (1982; 1992a). As logistical camps, these sites would represent locales used for
specialized resource procurement purposes. There are 19 LW sites in the intermediate diversity
category and they are spatially dispersed across the eastern UP. Most of them (n=13) are located
along Great Lakes shorelines, but there are six situated in the interior. Intermediate diversity

79

100
90
80
70
60

Floral

50

Faunal

40

Features

30
20
10
0
Extended

Intermediate

Limited

Figure 22: Percentage of sites by category with floral remains, faunal remains, or
subsurface features.

sites are less likely to include floral and fauna assemblages, but have a similar chance to include
features as extended diversity sites (Figure 22).
The limited diversity sites have the fewest activities and probably represent logistical
camps used by smaller groups for limited periods of time, for specific resource procurement
activities. The sites in the limited diversity category are well distributed across the eastern UP
and a much higher proportion of sites are located in the interior (18 of the 48 sites [37.5 percent])
than in the preceding categories. The most striking evidence for concerning the age of limited
diversity sites is that the sites that only included early LW components (n=3) were located on the
Great Lakes coastline, whereas seven of the eight dated interior sites were late LW in age and the
other site included both early and late LW materials. This is suggestive that use of the interior
was more prevalent in the late LW than in the early LW. Limited diversity sites are less likely to
include subsistence remains or features (Figure 22).

80

The diversity index appears to have established a framework in which to consider LW
sites in the eastern UP. This framework supports the contention that LW sites were used
differently. There is also evidence for spatial and temporal patterning with larger, extended
diversity sites being situated in coastal settings and interior settings only including intermediate
and limited diversity sites. It also appears that interior sites are more likely to have been used in
the late LW period. In the next chapter, Chapter 4, the spatial distribution of sites will be
considered in relation to a series of environmental variables to further assess LW settlement and
subsistence in the eastern UP.

81

4.0

LATE WOODLAND ARCHAEOLOGICAL SITE PREDICTIVE MODEL
This chapter will outline the development of a Geographic Information System (GIS)

predictive model of LW site locations on the Hiawatha National Forest (HNF). GIS is defined
as a set of computer tools for collecting, attributing, storing, transforming, and displaying spatial
data (Burrough and McDonnell 1998). The goal of this exercise was to identify settings or
locations of greater archaeological sensitivity, especially in regard to the environmental setting
of the LW sites. The ecological settings could be used to predict the potential occurrence of LW
archaeological sites in other parts of the Forest as well as the eastern UP of Michigan.
Archaeological surveys have been conducted on the HNF since the late 1970s leading to
the discovery of over 3,000 archaeological sites. Despite the relatively high number of
archaeological sites, there is often little specific information about the sites aside from location
and a limited assessment of age. Five hundred and seventy eight of the sites discovered on the
HNF are solely prehistoric or have a prehistoric component. Forty-eight of these can be
attributed to the LW period or include a LW component (Appendix J). The HNF sample
includes 59 percent of the LW sites in the eastern UP. In many cases the identification of a LW
component in the HNF or the eastern UP is based on the presence of diagnostic projectile points
and/or ceramics. Thus, we have archaeological site locations without much corresponding
information on specific resources that were being used at these sites.
The HNF includes approximately 898,980 acres (ac) (363,804 hectares [ha]) of land, in
two large swaths across the eastern UP (about 20 percent of the eastern UP land base) (Figure
23). These lands extend from Lakes Michigan and Huron in the south to Lake Superior in the
north and appear to serve as a viable proxy for the range of environments likely to be
encountered in

82

Figure 23: Hiawatha National Forest.
the eastern UP. The HNF is divided into an East Unit (EU) of approximately 396,943 ac
(160,637 ha) and a West Unit (WU) that includes some 502,037 ac (203,167 ha). These figures
reflect lands owned by the HNF, whereas the areal boundaries of the forest are larger including
approximately 1,298,205 ac (525,589 ha).

4.1

Development of the Model
The study of regional settlement patterns has a long history in archaeology (Flannery

1968; Trigger 1968; Willey 1953). Early on it was apparent that the distribution of sites on the
landscape could provide insight into how the sites were used and how the sites might interact
with one another. An outgrowth of this was the idea that the location of archaeological sites

83

could be predicted based on the attributes of known archaeological sites (Jochim 1976; Kohler
and Parker 1986). The advent of GIS technology and its application to understanding regional
settlement patterns has streamlined the process of integrating spatial, environmental, and
archaeological data (see Westcott 2000).
GIS-based archaeological modeling is the process of comparing geospatial data to other
variables and to forecast where people from the past used the landscape with sufficient intensity
to leave an archaeological signature (Dalla Bona 2000; Ebert 2004; Wescott 2000). It has been
defined as, “a technique to predict, at a minimum, the location of archaeological sites or
materials in a region, based either on the observed pattern in a sample or on assumptions about
human behavior” (Kohler and Parker 1986:400). The most prominent factors used to create
these models are environmental variables, such as the distance from a site to a particular resource
or the topography within a site location.
Predictive models make certain assumptions in order to operate. These assumptions are
quantified, compiled, and assessed as part of the modeling process. The first assumption in this
model is that the locations of known LW archaeological sites are representative of all the
existing LW sites in the HNF. A second assumption is that the LW sites are/were located in
relation to particular geographic or ecological features (e.g., topography, soils, distance to water,
vegetation, etc.). The second assumption is based on the expectation that the locations of LW
archaeological sites are correlated to the location of resources important to LW people.
For the current study, it was hypothesized that Late Woodland site settings, or locations
of increased LW archaeological sensitivity, could be predicted based on the environmental
settings of the LW sites already located on the Hiawatha. This sort of archaeological modeling
has been characterized as an inductive approach (Ebert 2004; Gibbon 2002; referred to as

84

empiric-correlative approach by Kohler and Parker 1986). This approach assumes that noncultural characteristics of a site location, in this case environmental characteristics, are useful
predictors of site location. In other words, we are relying on currently available information
about LW archaeological site locations on the HNF to develop generalizations or predictions
about other, currently unknown LW site locations on the HNF. Inductive models are the most
commonly used in archaeology (Ebert 2004; Gibbon 2002; Kohler and Parker 1986; Kwamme
1985). The other type of model, deductive models, rely on general theory concerning past
human behavior to hypothesize the locations of archaeological sites. Deductive models are less
often used in archaeology because they are more often based on more general and subjective
criteria as well as being more difficult to operationalize (Ebert 2004; Kohler and Parker 1986).
The existing model for LW settlement and subsistence in the eastern UP was derived
from a relatively small number of archaeological sites (as many as 29) located on the Great
Lakes shoreline and emphasizes the reliance on spring and fall spawning fish (the Inland Shore
Fishery model [Cleland 1982; see also Martin 1985; B. A. Smith 2004; Chapter 2.2]). The HNF
archaeological surveys have found coastal as well as interior LW sites adding an important new
dimension to our understanding of eastern UP LW settlement patterns (Dunham 2002; Dunham
and Branstner 1998).
More recently, pilot studies have shown that LW peoples used certain site settings and
habitats more extensively than others, including a greater use of interior locales than previously
expected (Dunham 2002, 2008; 2009; Franzen 1986, 1987; Martin 1999). While it can be
assumed that the distribution of LW sites reflects the location of resources used by LW peoples,
the distribution of sites is not entirely random and suggests other, likely cultural, factors play a
role in the selection of site locations.

85

Based on the results of the pilot studies, a more detailed analysis of LW site location on
the HNF was carried out and a GIS predictive model of LW site location on the Forest was
created (Figure 24). The goal of this exercise was to identify settings or locations of greater
archaeological sensitivity, especially in regard to the ecological setting of the LW sites. The
environmental settings were then used to predict the potential occurrence of LW archaeological
sites in other parts of the Forest as well as the eastern UP as a whole.
If we assume that archaeological sites are not randomly placed on the landscape, and that
their placement is an outcome of decision making by the people who used them, then the spatial
patterning should be observable (Ebert 2004; Gibbon 2002; Kvamme 1985; Warren and Asch
2000). A set of spatial and environmental variables were selected to better understand the
parameters of LW locations in the HNF. These variables -- aspect, distance to water, slope,
elevation, habitat classification, pre-1800 forest, potential growing days, and distance to potential
wild rice patches -- were considered to have potential predictive value based on our knowledge
of site distributions and LW adaptations.
There are 48 archaeological site locations with LW components that are included in the
HNF site files (Figure 25). Thirty-six of the LW sites on the HNF are found in the WU and the
other 12 in the EU. Most of these sites have been discovered through archaeological (cultural
resource) surveys on the HNF, although a small number have been found through other survey
efforts (Appendix J). Not only were they discovered through HNF surveys, they were found
using relatively consistent survey methods (Appendix K). For the purpose of this model, these
LW sites are considered a representative sample of the range of LW sites on the HNF.

86

Figure 24: GIS Model Flowchart.

87

Figure 25: Late Woodland Site Locations on the Hiawatha National Forest.

This a small sample of LW sites, but it represents the entire set of known LW sites on the
HNF from which to explore environmental setting. The use of modern environmental variables
may raise concerns. On the one hand, these can be considered proxy variables and that LW site
locations may co-occur with these variables, rather than correlate with them. However, the
major environmental variables explored in the study have currency in the Late Woodland period.
For example, forest communities similar to those of the present were established in the Upper
Great Lakes region by 3000 years ago and modern levels of the Upper Great Lakes had been
generally achieved by 2000 years ago (Anderton 1993; Brugam et al. 1997; Winkler et al 1986).

88

The original plan was to carry out a logistic regression analysis in the GIS to build a
model of the environmental variables that influenced Late Woodland site locations (Ebert 2004;
Warren and Asch 2000). However, the spatial scale of the HNF (898,980 ac) and small number
of LW sites (48) made such an analysis untenable. Thus, a decision was made to use the West
Unit LW site sample as the baseline for the construction of the model because of the higher
density of LW sites (502,037 ac and 36 sites) as well as the contiguous nature of the land base.
Unfortunately, a logistic regression analysis in the GIS for LW sites in the WU also
proved impractical. Although such an analysis could have been conducted, the spatial scales
would not have been fine enough for meaningful results (105 square kilometer quadrats [see
Chapter 4.1]). These constraints led to the development of a ranked model of site sensitivity
(Della Bona 2000; Ebert 2004).
There are an additional 33 LW sites in the eastern UP that are not located on HNF lands,
but are listed in the archaeological site files maintained at the SHPO. Five of these fall within
the aerial boundaries of the HNF, but are located on state land, private land or, in one case, the
Pictured Rocks National Lakeshore (Appendix J). The sites not listed in the HNF files have been
identified in a variety of ways including formal survey, informant sources, etc. Although these
sites, as well as the twelve LW sites on the EU, were not used as part of the model building, the
environmental variables derived from them were used in other aspects of this analysis. The
sample of LW sites from outside the HNF was not collected in as consistent a fashion as the
HNF sample, however it is still useful in that it supplies LW site data from additional site
locations to the study.
The spatial and environmental data were derived from GIS shapefiles and each of the LW
sites was plotted as a point with GIS. The spatial and environmental data pertinent to each site

89

locale were then generated for each archaeological site via GIS. Each of the LW sites was
entered as a point into the GIS, using Universal Transverse Mercator (UTM) coordinates for the
center point of the site. This rendered the site location as a 30 meter (m) × 30 m raster. A 30 ×
30 m raster covers a 900 m² area. While this is not a large area, it is the same size as the median
area of the LW sites in the eastern UP. Thus, half of the sites are larger and the other half are
smaller than the GIS-generated raster used in our study. The median site area in the WU sample
is 650 m². Coastal sites on the WU are a little larger with a median area of 1,000 m² and interior
sites are smaller at 625 m².
4.1.1 Distance to Water
Most analyses of prehistoric site location accept that access to water sources is a critical
factor in archaeological site location, and proximity to water is understood as an important
variable in northern Michigan and Upper Midwest (Ebert 2004; Franzen 1986; Gibbon 2002;
Kvamme 1985; Martin 1977; Peters 1986). Using the Michigan Geographic Data Library
(MiGDL) line and polygon hydrography datasets for lakes, ponds, rivers and streams, a
Euclidean distance raster was created at a 30 m × 30 m resolution (MiDGL 2013a, 2013b,
2013c). Using this dataset, the distance to water value was extracted to the LW site dataset,
allowing for the calculation of the distance to water. The resolution created a continuous scale
measurement in 30 m × 30 m increments (0 m, 30 m, 60 m, 90 m, etc.), but because these units
are square, some distances were calculated along the hypotenuse of multiple squares (0, 42.43,
67.08, 90.87, etc.). Regardless, the outcome simply describes a distance to water value. A site
listed as 0 m is directly adjacent to the source of water (the raster intersects the water), a site
listed as 30 m is within 30 m of a water source, etc. This method allows for the use of the output
as either continuous or categorical data.

90

4.1.2 Elevation
In northern Michigan, elevation has often been used as a guide in determining the relative
age of sites (Archaic or Woodland), specifically in relation to former and current Great Lakes
shorelines (Anderton 1995; Anderton et al. 2009; Lovis et al. 2012). The general premise is that
site locations varied with the past elevations of shorelines. Although elevation is not assumed to
be a reliable gauge of the relative age of coastal sites in the current study (diagnostic artifacts and
chronometric dating provide more secure dating), this variable may provide insights into broader
land use. The elevation of the archaeological sites was also generated through the MiGDL
Digital Elevation Model (DEM) (MiGDL 2013d). The elevation of each site locale is
documented in feet above mean sea level (amsl) based on the maximum elevation.
4.1.3 Slope
The slope, or steepness of ground, associated with a site is also a commonly used
environmental variable in archaeological site location models (Ebert 2004; Gibbon 2002;
Kvamme 1985, 1992; Martin 1977). Most models for archaeological site location assume that
residential sites are typically on level ground as opposed to steeply sloped areas, although
specialized activities and habitation locales may be situated in steeply sloped areas. The slope of
a given site locale in degrees was generated using the MiGDL DEM (MiGDL 2013d). Slope is
also based on a 30 m × 30 m resolution in the DEM and the steepest element represents the slope
of the raster.
4.1.4 Aspect
The aspect, or exposure, of a given locale has been considered an important variable in
archaeological site location (Ebert 2004; Franzen 1986; Gibbon 2002; Kvamme 1985). It is
understood that a southerly exposure in northern Michigan offers greater potential for warmth,

91

whereas more northerly and westerly exposures would be more susceptible to cold and
potentially more affected by prevailing winds (from the west-northwest) (Franzen 1986). The
aspect of each site locale was generated using the MiGDL DEM (MiGDL 2013d). The output
was presented in compass degrees (0.0-359.9 degrees). Site locations in level areas without a
discernible exposure were not included (4 sites) in the statistical analysis because they could not
be coded as 0.0-359.9 degrees. These sites are coded as -1 degree in Appendix L.
4.1.5 Growing Days
The adoption of maize horticulture is an important topic in the discussion of LW
subsistence in the Midwest (Brashler et al. 2000; Cleland 1983; Hart and Lovis 2013; O’Shea
2003; McElrath et al. 2000). Prehistoric maize is typically assumed to require 140 frost free days
to produce a reliable subsistence crop, though it does mature in a shorter period (Demeritt 1991;
Hart and Lovis 2013; Yarnell 1964). Microclimates along the Great Lakes shoreline can extend
the growing season, and in areas with a growing season of 120 to 140 days, maize is certainly
possible (O’Shea 2003). Summer heat, or growing degree days (GDD), is also a factor (Demeritt
1991).
A shapefile was created in the GIS that identifies the growing season (frost free days or
growing days) for the eastern UP (Figure 26). The shapefile was created through the
georectification of maps from Eichenlaub et al.’s (1990) climate atlas, depicting growing days in
increments of ten days (≤ 100, 100-109, 110-119, 120-129, 130-139, 140-149). This map is a
depiction of frost free days, but does not incorporate growing degree days into its zones. The
site locations were integrated within the shapefile and the estimated growing days for each site
were extrapolated The growing days variable, along with aspect, may provide an insight into the
placement of sites if potential maize horticulture was a factor. Each of the LW sites (20CH6,

92

20MK1, and 20MK24) that have produced maize remains are situated in the 140+ day zone (see
Chapter 3.5).
4.1.6 Distance to Potential Wild Rice Stands
Wild rice is not generally associated with Native American subsistence in the central or
eastern portion of Michigan’s UP, despite its well documented use in Wisconsin, Minnesota, and
Ontario (Brashler et al. 2000; Cleland 1983; Jenks 1900; Jenness 1935; Vennum 1988). It is
presently unknown to what extent LW cultures in the HNF area may have relied on wild rice
gathering. Despite this, there are at least 40 locations in the eastern UP where wild rice can be
documented in the nineteenth and twentieth centuries (Dunham 2008). The locations of these

Figure 26: Growing Days.

93

wild rice locales correlate with the distribution of LW sites based on a previously completed
pilot study (Dunham 2008)
The locations where wild rice was documented were derived primarily from early
twentieth-century United States Forest Service and Michigan Department of Conservation
studies (HNF 1938; MDNR c. 1949; Miller 1943; Pirnie 1935); university herbarium collections
(Edman 1969; MSU Herbarium nd); and historical sources (especially Jenks 1900; Schoolcraft
1966:201; 1975:115). Further refinement of the locales was also gleaned from more recent
research conducted by the Great Lakes Indian Fish and Wildlife Commission and the Bay Mills
Chippewa Tribe (Brown 1999; Lu and Waller 1996).
A buffer was created around modern lake shoreline polygons in the GIS for locations
where wild rice was documented in the nineteenth or twentieth century (Figure 27). The buffer
was expanded at 100 m intervals to 1,500 m. The use of 1,500 m was somewhat arbitrary, but
based on the assumption that sites greater than 1,500 m from a potential wild rice patch were less
likely to have made significant use of that resource. The 1,500 m radius can be considered the
maximum distance one might travel from a residential camp to a wild rice processing or storage
locale. A 700 m radius was also considered, half of the 1500 m radius rounded down to the
nearest 100, as a more conservative distance for travel from a residential site to a wild rice patch.
Wild rice is a bulky item that ethnographically was processed and stored on near-shore
landforms overlooking the wild rice stand (Johnson 1969; Mather and Thompson 2000; Vennum
1988).

94

Figure 27: Location of Wild Rice Patches.

4.1.7 Pre-1800 Vegetation
Given the recentness of the LW period and the generalized stability of the UP
environment from the LW through the present, models of the pre-1800 vegetation can serve as
proxies for the vegetative communities likely present in the LW (Brugam et al. 1997; Dunham
2009; Winkler et al 1986). Several recent studies of pre-1800, or pre-European, settlement forest
composition in the eastern UP have been undertaken that have relied on data derived from
General Land Office (GLO) surveys (Comer et al. 1995; Delcourt and Delcourt 1996; Zhang et
al. 2000). The reconstruction of the forest types is based on the identification of witness and
bearing trees during the original land surveys between the 1820s and 1850s (Albert and Comer

95

2008; Delcourt and Delcourt 1996). The study by Comer et al. (1995) also digitized the data in a
GIS format (MiGDL 2013e).
The range of vegetation was grouped in the following categories and treated as a
categorical variable: unclassified; jack pine; mixed pine; mixed upland conifer forest (hemlock
dominated); northern hardwood; lowland conifer; lowland hardwood; and wetland/marsh. These
categories were broadly derived from the codes developed by Comer et al. (1995:9) as well as
from information presented in Kost et al. (2007) and Coffman et al. (1980). The categories
provide a continuum characterizing vegetation types from the driest (jack pine) to wettest
(wetland/marsh). Areas that were not attributed in Comer et al. (1995) are coded as unclassified
(Appendix M).

4.1.8 Habitat Classification System
The GLO based pre-1800 vegetation models work well to determine an aggregate sample
of forest composition, but is not well suited to identify specific patches or stands on the
landscape. The Habitat Classification System is based on the understanding that plants are
usually found in predictable patterns or communities that are often associated with particular soil
types (Burger and Kotar 2003; Coffman et al. 1980; Kotar 1976). The habitat system is an
ecological model predicting the likely succession pattern and climax species in a given niche,
whereas the GLO derived data is historical, a snapshot of the forests as the existed in the midnineteenth century. The benefit of the habitat data is that it provides a more fine-grained setting
for the site itself, whereas the GLO data provides a coarser-grained view of the broader trends in
forest composition (Dunham 2009).

96

The soils in the eastern UP have been classified using the Habitat Classification System
and the soils data are available in GIS (MiGDL 2013f; Coffman et al. 1980). This data set is a
digital soil survey and generally is the most detailed level of soil data available. The Habitat
Classification System has shown great potential in previous studies of UP prehistoric
archaeology (Dunham 2009; Franzen 1986). There is a reasonable expectation that modern soils
reflect the general conditions that were prevalent in the LW (see Dunham 2009). Like the pre1800 vegetation, the range of habitats was grouped in the following categories: unclassified;
jack pine; mixed pine; mixed upland conifer forest (hemlock dominated); northern hardwood;
lowland conifer; lowland hardwood; and wetland/marsh (Appendix N).

4.1.9 Site Setting
Interpretations of Late Woodland settlement suggest that site function and seasonality
varied with location relative to the Great Lakes shore. This study considers two broad
environmental settings in relation to each site: Great Lakes Coastal (or simply Coastal) and
Interior. This simply refers to the site being located along or near one of the Great Lakes
(Michigan, Huron, or Superior) or in the interior of the eastern UP. This variable was more
subjectively assessed and not determined through the GIS because the diversity and nature of
Great Lakes littoral settings could not be simply quantified as a distance from the shore of a
modern Great Lake (Albert 2003; Kost et al. 2007; Lovis et al. 2012). Each site was coded and
the variable added to the analysis. The site setting variable has shown potential for exploring
changes in settlement and subsistence strategies (Dunham 2002; 2008; 2009).

97

4.1.10 Random Points
In addition to the 36 LW site locations on the WU of the HNF, 50 locations were also
randomly generated across the WU (random points) within the GIS (Figure 28). The same
spatial and environmental variables considered for the archaeological sites were also generated
for each of the random points. This allowed for the comparison of known LW site locations and
associated environmental attributes, with those of a randomly generated dataset of broadly
comparable size. That is, the attributes for the 50 random sites could be considered
representative of the total study area, and this allowed for a statistical comparison of attributes of
site-bearing locations and attributes of the generalized study area.

Figure 28: Random Points on the West Unit of the Hiawatha National Forest.

98

4.2

Statistical Analysis
The exploration of the distribution of LW and random points was begun by carrying out

summary statistics for each set of data and for each variable (Appendix J). Each of the
continuous variables was compared to one another with Pearson’s correlation coefficient (Table
5). None of the variables had a statistically significant correlation with one another which
indicates their independence and permits model building. Three primary statistical tests were
used to analyze the data: Welch’s Approximate t-Test, Kolmogorov-Smirnov Two Sample Test,
and Chi-Square Test of Independence.
Variable

Slope

D. to Water

Elevation

Aspect

Growing Days

-

-0.010

-0.028

0.335

0.046

D. to Water

-0.010

-

0.106

-0.084

-0.087

Elevation

-0.028

0.106

-

0.130

-0.406

Aspect

0.335

-0.084

0.130

-

-0.173

Growing Days

0.046

-0.087

-0.406

-0.173

-

Slope

Table 5: Pearson’s Correlation Coefficient.
4.2.1 t-Test
Welch’s approximate t-Test was used to compare archaeological site and random site
locations where the variables were continuous (distance to water, slope, elevation, aspect, and
growing days) (Sokal and Rohlf 1995:404-406). Welch’s t-Test is a useful mechanism to
compare the means of two samples in which the variances are assumed to be different. This
experiment tests the hypothesis that archaeological sites and random points are similarly
distributed around the sample mean (H0) or are differently distributed (H1).

99

The results of the t-Tests comparing LW sites and random points are presented in Table
6. These tests were each conducted at a 0.1 confidence level. The null hypothesis was rejected
for three variables (distance to water, elevation, and slope), and was not rejected for either aspect
or growing days. The results of the tests of the aspect and growing day variables suggest there is
not a significant difference in the distribution of LW sites or random points on the landscape and
that these variables may be less important in archaeological site location selection. The
comparison of the mean distance to water between these LW and random points shows that LW
sites are situated more closely to water than random points. Likewise, LW sites are distributed at
lower elevations than random points. The slope variable test indicates that the random points are
more likely to be located in areas with less degree of slope than LW sites.
LW Sites v.
Random Points
Distance to Water
Slope
Elevation
Aspect
Growing Days

Observed t Statistic

Critical Value @
0.1 (two-tailed)

P Value
(two-tailed)

Statistically
Significant

7.530

1.674

0

Yes

-2.291

1.681

0.027

Yes

5.008

1.665

0

Yes

1.150

1.666

0.254

No

-0.428

1.669

0.679

No

Table 6: Summary of t-Tests.
4.2.2 Kolmogorov-Smirnov Test
The Kolmogorov-Smirnov two sample test (KS test) was used for the continuous
variables as well as the categorical variables for pre-1800 vegetation, habitat, and setting. This
test compares the difference between two distributions. The null hypothesis (H0) assumes that
the two samples are dispersed identically. While the KS test is considered more appropriate for
continuous variables (Sokol and Rohlf 1995:434-439), recent studies have used this test for

100

categorically arranged data (Ebert 2004; Thompson and Turck 2009; Whitcomb et al. 2002). It
is considered a conservative test when used with categorical data.
The results of the KS tests comparing LW sites and random points are presented in Table
7. These tests were each conducted at a 0.1 confidence level. The results of the tests on
continuous data were largely comparable to the t-Tests with the null hypothesis rejected for
distance to water, elevation, and slope and were not rejected for either aspect or growing days.
When the categorical data were explored through KS-tests, pre-1800 forest appeared to have
similar distributions for LW sites and random points, whereas habitat and setting
(coastal/interior) did not.
LW Sites v.
Random Points

Observed KS
D Statistic

Critical Value
@ 0.1

P Value
(two-tailed)

Statistically
Significant

Distance to Water

0.817

0.267

0

Yes

Slope

0.291

0.267

0.054

Yes

Elevation

0.531

0.267

0

Yes

Aspect

0.270

0.119

0.285

No

Growing Days

0.166

0.267

0.586

No

Habitat

0.574

0.267

0

Yes

Pre-1800 Forest

0.260

0.267

0.111

No

Coastal/Interior

0.452

0.267

0

Yes

Table 7: Summary of Kolmogorov-Smirnoff (KS) Test of Site Frequency Distribution.
KS-tests were also carried out for the habitat and pre-1800 forest variables comparing
them with the distributions of habitat types and pre-1800 forest types for the WU of the HNF
(Tables 8 and 9). The proportion of each variable for the WU as a whole was calculated through
GIS. The distribution of random points was not statistically different than the proportion of
habitats and pre-1800 forest types in the WU. The distribution of LW sites and pre-1800 forest

101

types was not statistically significant. A comparison of LW sites and WU habitats did reject the
null hypothesis.
Random Points v.
West Unit

Observed K-S D
Statistic

Critical Value
@ 0.1

Critical Value
@ 0.2

Statistically
Significant

Pre-1800 Forest

0.081

0.211

0.185

No

Habitat

0.090

0.211

0.185

No

Table 8: Summary of Kolmogorov-Smirnoff (KS) Test of Site Frequency Distribution.
LW sites v. West
Unit

Observed K-S D
Statistic

Critical Value
@ 0.1

Critical Value
@ 0.2

Statistically
Significant

Pre-1800 Forest

0.210

0.237

0.208

No

Habitat

0.574

0.237

0.208

Yes

Table 9: Summary of Kolmogorov-Smirnoff (KS) Tests.
4.2.3 Chi-Square test
The chi-Square (χ²) test of independence is employed to test the difference between an
actual sample and another, hypothetical or previously established distribution. In this case, the
chi-square test was used to test the differences between the distributions of LW sites and random
points. Previously completed pilot studies relied on chi-square tests to explore the differences in
site distributions (Dunham 2008; 2009). The null hypothesis (H0) assumes that the two samples
share a common distribution. Chi-square tests were used to assess the distance to water variable,
distance to wild rice variable, the site setting variable, and the habitat variable because the results
of the t-Tests and KS-tests showed these to be the most critical variables.
In addition to the comparison of WU LW sites and random points, the results of chisquare tests comparing all the LW sites in the eastern UP (81 LW sites) and a separate set of
eastern UP random points (80 random points) are also presented on Table 10. This was done to
provide additional comparative data for the generation of the predictive model because of the
small sample of LW sites.

102

Observed χ²
Statistic

Critical Value @
0.1

P Value

Statistically
Significant

D. to Water (EUP)

121.75

4.605

0

Yes

D. to Water (WU)

48.118

4.605

0

Yes

4.596

2.706

0.032

Yes

0.803

2.706

0.370

No

7.174

2.706

0.007

Yes

1.531

2.706

0.216

No

Habitat (EUP)

16.930

2.706

0

Yes

Habitat (WU)

30.071

2.706

0

Yes

64.571

2.706

0

Yes

23.203

2.706

0

Yes

LW Sites v.
Random Points

D. to Wild Rice
1500 m (EUP)
D. to Wild Rice
1500 m (WU)
D. to Wild Rice
700 m (EUP)
D. to Wild Rice
700 m (WU)

Coastal/Interior
(EUP)
Coastal/Interior
(WU)

Table 10: Summary of Chi-Square Tests.
As with the KS-test, the chi-square test of the site setting variable confirmed that there is
a significantly greater correlation between LW sites and coastal settings than there is between
random points and coastal settings. Nearly half (47.2 percent [17 of 36]) of the WU LW sites are
in coastal settings and only one of the 50 random points (2 percent) is in a coastal setting. For
the eastern UP as a whole, 69.1 percent (56 of 81) of the LW sites are in coastal settings and only
7.5 percent (6 of 80) of the random points are in coastal settings.
The distance to water variable was set as three categories: 0-119 m; 120-239 m; and 240+
m. Thirty three of 36 WU LW sites (91.7 percent) and 87 percent (71 of 81) of all the sites in the
eastern UP are located within 120 m of a source of water. Conversely, 66 percent (30 of 50) of
the WU random points and 87.5 percent (70 of 80) of all the eastern UP random points are
greater than 240 m from water. Based on these tests, an overwhelming number of LW sites can
be expected to be found within 120 m of a source of water. Most of the random points are

103

located over 240 m from a source of water. The results of these tests, along with the t-Tests and
KS-tests, demonstrate that LW sites have been most frequently encountered within 120 m of
water.
The KS-tests on the habitat variable demonstrated that LW sites have been more
frequently found in mixed pine habits than other habitats. This fact, and to allow for a valid chisquare test, led to a comparison between LW sites and mixed pine habitat v. other habitats for the
WU and for the eastern UP. The results of both these tests are statistically significant reflecting a
greater use of mixed pine habitats by LW people than expected. About two-thirds (24 of 36) of
the WU LW sites are situated in mixed pine habitats, whereas only about 12 percent of WU
habitats are mixed pine habitats. For the eastern UP as a whole, 39.5 percent (32 of 81) of the
LW sites are in mixed pine habitats and these habitats comprise only 8.4 percent of the eastern
UP.
The distance to wild rice variable was initially derived from a comparison of LW sites
and the eastern UP as a whole. About 14.8 percent (12 of 81) of LW sites are located within
1,500 m of a twentieth century wild rice patch and about 13.6 percent (11 of 81) of them are
within 700 m of one. Although this does not seem significant, only one (1.25 percent) of the
eastern UP random points falls within 700 m of a wild rice locale and only 3 (3.8 percent) fall
within 1,500 m of one. The chi-square test shows both to be significant results. When the same
variables are compared on the WU, the proportion of random points increased to 6 percent (3 of
50) within 700 m and to 8 percent (4 of 50) within 1,500 m. Despite a proportional increase of
LW sites to 16.7 percent (6 of 36) within 700 m and 1500 m on the WU, the chi-square test does
not reject the null hypothesis.

104

4.3

Quadrat Analysis
A spatial overview of the WU LW sites was performed using quadrat tests with a 1000 ×

1000 m (1 square kilometer [km]) quadrat as the basic unit of analysis. The decision to use a 1
km² unit was made after calculating the optimal size of a quadrat and coming up with a figure of
nearly 105 square km (Quadrat Size = 2A/r; where A = the area of the study area [1,888.44 km²]
and r = the number of points in the distribution [36 LW sites on the HNF WU] [Wong and Lee
2005]). While 105 km² is a similar size to the quadrats used in the pilot studies (36 mi² [93.2
km²] [Dunham 2008; 2009]), it is too coarse of a scale for meaningful analysis for the purposes
of generating a predictive model for archaeological sites.
The 1 km² quadrats were placed over the WU and focused on lands owned by the HNF
(the boundaries of the WU include state, county, private, and other federal lands). There are
2641 one kilometer square quadrats that comprise this area (Figure 29). Some of the quadrats
extend beyond the HNF owned lands as well as onto the Great Lakes which leads to an increased
number of quadrats. The finer scale quadrats are also problematic because archaeological sites
are only present in 28 of them (about 1 percent of the quadrats), but it was hoped the finer scale
would be more useful for identifying meaningful trends in the environmental data.
Thirty five LW sites fall on HNF lands in the WU. One site, 20DE106, is located on
private land, but listed in the HNF files (FS 09-10-01-076). This site was used as part of the
statistical analyses, but is not part of the quadrat analysis because it did not fall within a quadrat.
The 35 remaining LW sites on the WU of the HNF fall within 28 quadrats (Figure 29). One
quadrat includes four sites, one includes three, and two include two sites. The quadrats with four
and three sites, as well as one of the quadrats with two sites are located on Grand Island. Seven

105

Figure 29: West Unit Quadrats.

106

of the sites are located on Murray Bay on the south side of the island and the other two at the
mouth of Echo Lake Creek on the west side of the island. The other quadrat with two sites is
situated on the lower Sturgeon River upstream from Nahma. The remaining 24 quadrats include
a single LW site.
The fifty random points generated for the WU were also examined and each is situated
within a single quadrat. One of these quadrats includes a LW site as well as a random point.
Two thousand five hundred and sixty four quadrats do not include either LW sites or random
points.
While the spatial outcomes of the quadrat tests were of somewhat limited value, a review
of the environmental variables was quite enlightening. For the most part the quadrat data
provides a generalized background for the WU of the HNF that can provide part of the baseline
for these analyses. Beginning with elevation, the pooled means of the 2641 quadrats generated a
mean elevation of 751.3 ft (median 753.6 ft). This includes a range of elevations from 581 ft
(the elevation of Lake Michigan/Huron), to a maximum elevation of 1079 ft. This compares
well to the mean elevation of the random points, but not to the LW sites (Table 11).
As shown in Table 11, the LW sites are at a lower mean elevation than the quadrats or the
random points. This is because about half of the WU LW sites are situated on or near the shores
of the Great Lakes (ca. 582 ft for Lakes Michigan and Huron and ca. 602 ft for Lake Superior).
The relatively low elevation of the Great Lakes shorelines in relation to the eastern UP as a
whole skews the mean elevation of the LW sites to a lower elevation. There is a relationship
between site location and Great Lakes Coastal locations, but this relationship can be subsumed
under the occurrence of LW sites and all sources of water (e.g., distance to water).

107

Elevation

All Quadrats

Random Points

LW Sites

Mean

758

748

658

Median

759

752

619

Minimum

581

594

582

Maximum

1079

921

856

Table 11: Elevation in Feet amsl.
When slope is considered (Table 12), the slope of the pooled mean of the quadrats and
the mean slope of the random points were low (6 and 8 degrees respectively), whereas the mean
slope of the archaeological sites was higher (14 degrees). The quadrat data shows that there is
not a lot of topographical relief in the WU and the ground surfaces are generally relatively level.
The explanation for the higher elevations of LW sites is more fully detailed below, but the basic
rationale relates to the relationship between site locations and bodies of water.
Slope

All Quadrats

Random Points

LW Sites

Mean

8

6

14

Median

6

3

6

Minimum

0

0

0

Maximum

85

35

82

Table 12: Slope in Degrees.
The aspects of LW sites and random points were not statistically different from one
another. The quadrat test provides support for the conclusion that the lands in the WU, and the
eastern UP in general, are oriented in a southeasterly manner (Table 13). The LW sites have a
mean aspect of 163 degrees (median 165 degrees), the random points have a mean of 160
degrees (median 161 degrees), and the pooled mean of the quadrats is 165 degrees (median 154
degrees). If we assume a southerly aspect is between 120 degrees and 240 degrees, then the
mean of all three sets fall comfortably within that range. This suggests that southerly, and

108

Aspect

All Quadrats

Random Points

LW Sites

Mean

165

160

163

Median

154

161

165

Table 13: Aspect in Degrees.
specifically, southeasterly (120 to 180 degrees), aspects are well represented in the eastern UP.
Aspect does not appear to be a critical variable in predicting site location in the region.
The growing days variable was also examined. The pooled mean growing days of the
quadrats as well as the means of the random points and LW sites were each 115 days or less
(Table 14). The results, along with the previous statistical studies, suggest that LW site locations
were not solely selected based on the potential to carry out horticulture (minimally 120 growing
days, but more likely 140+ growing days). The quadrat data informs us that areas with optimal
growing days for maize are uncommon in the eastern UP. Of the 2641 quadrats in the WU, only
173 include areas with 140+ growing days (6.5 percent).
Growing Days

All Quadrats

Random Points

LW Sites

Mean

115

113

114

Median

110

110

110

Table 14: Growing Days.
The quadrat approach is more effective in discussing some of the other environmental
variables, such as the relationship between LW sites, water sources, and habitats. As previously
discussed, the distance to a source of water is a critical variable in determining LW site locations.
Quadrats that include LW sites have a mean minimum distance to water of 63 m, whereas
quadrats with random points, and all the WU quadrats, have a mean minimum distance of greater
than 300 m to water (Table 15).

109

Mean Distance to Water
Habitat Diversity

All Quadrats

Random Points

LW Sites

444.16

316.78

62.50

3.1

3.1

3.6

Table 15: Mean Distance to Water and Habitat Diversity.
The habitat variable provides another measure of assessing site selection on the
landscape. We have seen that LW sites are most commonly associated with mixed pine habitats.
The quadrat study confirms this observation and also illustrates that LW peoples were selecting
areas with greater habitat diversity than other areas (Table 15). Quadrats that include
archaeological sites have a density of 3.6 habitats per quadrat, whereas the quadrats with random
points as well as for all the quadrats, have a diversity of 3.1 habitats per quadrat. Mixed pine,
northern hardwood, lowland conifer, and open wetland habitats were each present in over 70
percent of quadrats with LW sites (Figure 30). Only northern hardwood and lowland conifer
were present in over 70 percent of the quadrats with random points as well as the entire set of
WU quadrats. Northern hardwood and lowland conifer habitats are the best represented habitats
in the WU and in the eastern UP as a whole (WU 37.7 and 36.8 percent respectively, EUP 30.3
and 42.6 percent), whereas mixed pine and open wetland habitats are not (WU 12.1 and 3.6
percent respectively, EUP 8.4 and 7.6 percent).
Wild Rice patches appear in only 67 quadrats. For perspective, interior lakes appear in
998 quadrats. Using an a priori probability approach and the wild rice variable we predict that
wild rice would be found in 0.025 quadrats in the WU (67 of 2641 quadrats). LW Sites are
currently known in 28 quadrats, thus one would expect wild rice in less than 1 (0.7 quadrats) of
the quadrats with a LW site. Wild rice and LW sites appear in 5 quadrats which is a ratio of
about 7:1 observed to expected. A chi-square test confirms the statistical significance of this

110

Figure 30: Comparison of Quadrat Habitats.

difference (Table 16). For random points (50 quadrats), one would expect 1.25 quadrats to have
wild rice. Wild rice and random points appear in 3 quadrats (a ratio of 2.4:1). While the higher
than expected number of quadrats with both wild rice and random points appears significant, the
chi-square test does not confirm this interpretation (Table 12).
Wild Rice

Observed χ²
Statistic

Critical Value @
0.1

P Value

Statistically
Significant

LW Sites

6.776

2.706

0.009

Yes

0.377

2.706

0.539

No

Random Points

Table 16: Wild Rice in West Unit Quadrats (Yates Correction).

111

4.4

Review of the Variables
The following discussion examines the results of selected statistical tests, focusing on

those variables that were significant (distance to water, slope, elevation, habitat, site setting, and
wild rice).

4.4.1 Site Setting
The Inland Shore Fishery model assumes the importance of LW coastal sites for access to
spring and fall fishing grounds (Cleland 1982; Martin 1985). Therefore, one would expect a
high degree of coastal setting sites. A pilot study has shown that LW coastal sites outnumber
LW interior sites by a ratio of a little more than 2:1 for the eastern UP (Dunham 2002). The KStests and chi-square tests show a greater proportion of LW sites (observed) in coastal settings
than random points (expected). These tests confirm that coastal sites outnumber interior sites by
a little over 2:1 for the eastern UP as a whole, but also show that this ratio is slightly over 1:1
favoring interior sites for the WU of the HNF (19 interior to 17 coastal).
The first response to the difference in ratios is that the WU of the HNF has included a
significant amount of archaeological survey on interior areas, whereas many other parts of the
eastern UP have not. However, interior locales in the East Unit of the HNF have been similarly
surveyed and the ratio is 7 coastal LW sites to 1 interior LW site (14 coastal to 2 interior). A χ²
test comparing WU LW site settings and the settings of the remainder of the eastern UP LW sites
indicates that the pattern on the WU is distinct and statistically significant (Observed χ² = 14.583,
Critical Value @ 0.1 = 2.706, P Value = 0). This suggests that LW people are making greater
use of the interior in the WU than in other parts of the UP. Additional archaeological survey for

112

LW sites in the eastern UP, aside from the West and East Units of the HNF, may clarify this
difference.
Martin’s (1985) study of Woodland coastal settlement within the Upper Great Lakes,
with particular reference to the eastern UP and the Straits of Mackinac, found that site location
was most closely related to local fish habitat conditions, with most site locations displaying
complex shorelines (especially associated with embayments). Archaeological field work in the
HNF generally confirms these studies, associating prehistoric sites with former and modern
barriers, spits, embouchures and other complex landforms associated with embayments
(Anderton 1993, Anderton 1995; Dunham and Branstner 1998). The sensitivity of coastal
settings is further demonstrated through the observation that nearly 32 percent of 173 prehistoric
sites recorded and tested between 1992 and 1997 were associated with modern or
paleoshorelines (Dunham and Branstner 1998:165-167). Nearly all of these sites are associated
with complex landforms associated with embayments and river/stream mouths.

4.4.2 Distance to Water
By all measures, distance to water is a critical factor in LW site location in the eastern
UP. An overwhelming majority of LW sites (87 percent of all LW sites) have been found within
120 m of a source of water. Most of the random points (87.5 percent of all random points) are
located over 240 m from a source of water. The quadrat test also found that quadrats with LW
sites were closer to water than other quadrats. The results of these tests, along with the t-Tests
and KS-tests, demonstrate that known LW sites are located in close proximity to water.

113

4.4.3 Elevation
The statistically significant differences in elevation between the random points and LW
sites is the result of the high proportion of LW sites being located along or near a Great Lake
shoreline (ca. 582 ft for Lakes Michigan and Huron and ca. 602 ft for Lake Superior). The
relatively low elevation of the Great Lakes shorelines in relation to the eastern UP as a whole
skews the mean elevation of the LW sites to a lower elevation. This can be illustrated by
comparing the means of coastal and interior setting LW sites for the entire eastern UP. The
mean elevation of coastal setting LW sites is 602 ft (minimum elevation of 582 ft and maximum
of 630 ft). There is only a difference of 48 ft separating the highest and lowest elevation for
coastal setting LW sites and coastal sites comprise 69 percent of the LW sites in the eastern UP
sample. Interior LW sites have a mean elevation of 693 ft (minimum elevation of 583 ft and
maximum of 856 ft) and a spread of 303 ft from the highest to the lowest. Further, only one of
the 50 WU random points is located in a coastal setting which further effects the distribution of
elevations, since 49 of the random points are located in the interior.
Elevation is clearly a significant factor because of the high proportion of archaeological
sites situated in coastal settings. However, the distance to water is a more sensitive variable in
relation to site location and, by default, coastal settings are all closer to water than the WU
random points. The mean distance to water for random points is 459.4 m.

4.4.4 Slope
It seems counterintuitive that random points would exhibit less slope than LW sites. One
tends to associate level areas with habitation locales. However, when we consider how LW sites
are dispersed across the landscape, this can be explained by topography and the terms of

114

measurement in the GIS system. First, LW sites are typically located in close proximity to water
such as a Great Lake, an interior lake, or an interior river or stream. Each of these settings is
often associated with a bank or terrace that is situated above the body of water.
For example, the Williams Annex site (20AR353 [FS 09-10-03-811]) is situated at the
base of a steep Nipissing era wave cut bluff on a lower surface overlooking Lake Superior and
the Widewaters site (20AR245 [FS 09-10-03-667]) is situated on a terrace above the Indian
River (Franzen 2000; Robinson et al. 1991; Rutter and Weir 1985). The 30 m square raster used
to measure slope in the GIS takes the highest value within the 30 m square. The Nipissing bluff
is the basis of the 84 degree slope at the Williams Annex site and the terrace along the Indian
River is the cause of the 23 degree slope for the Widewaters site, despite the fact that each of
these sites is situated on relatively level ground. In other words, the LW sites are often located
on relatively level ground in areas with high locale relief (see also Kvamme 1992). For the
reasons enumerated above, slope, as measured in the GIS, is not considered a critical variable in
LW site location.

4.4.5 Habitat
Previously completed pilot studies have shown a statistically significant relationship
between the location of known LW archaeological sites in the eastern UP and mixed pine
habitats (Dunham 2009; 2012). Similarly, Franzen’s (1987) study found that most interior lake
sites were associated with pine-oak forests (generally equivalent to the mixed pine forest
category used by Dunham [2009; 2012]). The current study demonstrates that about 66.7 percent
of WU LW sites and 49.5 percent of LW sites in the eastern UP are located in mixed-pine
habitats. These habitats comprise about 12 percent of the WU and less than 10 percent of the

115

habitats in the eastern UP. The statistical tests presented above indicate that this is not the
expected distribution of sites and habitats and that there is a positive correlation between LW
sites and mixed pine habitats.
The mixed-pine habitats provide a variety of resources that were attractive to Woodland
peoples in the region. The succession pattern in these forests is conducive to beaver, moose, and
warm-season deer habitat because of the regular creation of openings, especially through fire,
and a high incidence of aspen and other habitats preferred by these herbivores (Dunham 2012;
Franzen 1986, 1987). Such habitats also include a higher incidence of certain fruits, such as
blueberries as well as other resources, such as acorns, that were utilized as food by Native
Americans as well as by the animals they hunted (Anderton 1999; Dunham 2000a; 2009; Chapter
3.5).

4.4.6 Wild Rice
A pilot study exploring the modern distribution of wild rice and the distribution of
Woodland period archaeological sites found a spatial relationship between the archaeological
sites and places where wild rice grows (Dunham 2008). Chi-square tests were employed in the
pilot study to compare the relationship between townships (93.2 km² quadrats) that have wild
rice and archaeological sites. The results of the chi-square test showed a statistically significant
relationship between townships that have wild rice and archaeological sites as well as for
townships that do not include either (Dunham 2008).
As part of the current study, a statistically significant relationship was found between LW
sites within 1500 m and within 700 m of a modern wild rice stand as opposed to random points
for the eastern UP as a whole. This relationship was also supported in the comparison of WU

116

LW site locations as part of the quadrat analysis. The data from the pilot study and the eastern
UP portion of the current study support the potential significance of the wild rice variable,
though the variable does not seem as important in LW site location dynamics as habitat or
distance to water.

4.5

Model Construction
The preceding statistical analyses indicated that distance to water and habitat were the

most sensitive variables in regard to WU LW site location. Distance to potential wild rice stands
was also found to have a statistically significant relationship with the location of LW sites, but
not across all levels of analysis. The distance to water variable and habitat variable were
explored through a multiple regression formula. The results of this test showed that the
combination of these two variables was more likely to predict LW site locations than either of
the variables individually (Appendix O). This supports the utility of the model.
A working framework was developed in which distance to water and habitat were ranked
and assigned point values. Categories were developed for each variable from the statistical tests:
distance to water - 0-120 m, 121-240 m, and 240+ m (the 120 m scale allowed for more efficient
processing in GIS); habitat – unclassified, jack pine, mixed pine, upland conifer, northern
hardwood, lowland conifer, lowland hardwood, and marsh/wetland; and the conservative
measure for wild rice – 0-700 m and 700+ m. The point values were generated by treating the
WU LW site set and the eastern UP LW site set as two distinct sets, though they overlap. The
two sets combined to include 117 LW sites. Mean scores were generated by assessing the
number of LW sites in each category and dividing by the total. These scores were then rounded
up or down to a whole number (Table 17). The wild rice score was determined as presence or

117

absence, with sites situated within 700 m of a patch receiving points and sites greater than 700 m
from a stand receiving no points. The distance to water, habitat, and wild rice scores were then
combined to generate a score for each site. In the same manner, scores were also generated from
random points.
The scores in Table 17 are values reflecting the currently understood occurrence of LW
sites within that category. For example, since approximately 90 percent of the LW sites have
been found within 120 m of a source of water, it is hypothesized that like areas will produce 90
percent of the LW sites to be discovered. Combining the scores from each of the three variables
provides a weighted score with the assumption that the higher the score, the more likely an area
is to include a LW site.
Locations within 120 m of a source of water and located in mixed pine habitats were
determined to represent the areas of highest potential for LW sites (Table 18). Combining the
distance to water score and habitat score, a score of 140 is achieved for these locations. If one
of these locations was also within 700 m of a wild rice stand, then the score would be increased
to 155. Medium sensitivity areas were defined by scores between 59 and 139. The base-line of
59 points is the value of a location in a mixed pine habitat between 120 m and 239 m of a source
of water. Low sensitivity areas have scores of less than 59 points.

4.6

Review of the Model
The archaeological predictive scores were then applied to the sample of LW sites to

observe the distribution of the scores in relation to the sites (Appendix P). When the scores are
compiled against the LW sites in the WU, 63.9 percent fall within the high category, 33.3 percent
in the medium category, and 2.8 percent in the low category. In other words, 97 percent of the

118

Variables

Score

Distance to Water
0-120 m

90

121-239 m

9

240+ m

1

Habitat
Unclassified

4

Jack Pine

8

Mixed Pine

50

Upland Conifer

2

Northern Hardwoods

15

Lowland Conifer

18

Lowland Hardwoods

0

Marsh/Wetland

3

Wild Rice
0-700 m

15

700+ m

0

Table 17: Late Woodland Site Predictive Model Point Scales.

Archaeological Sensitivity

Sensitivity Score

High

140+

Medium

59-139

Low

0-58

Table 18: Late Woodland Site Archaeological Predictive Scores.
known WU LW sites have been found in areas of either high or medium sensitivity for LW sites.
The WU random points fall in a nearly inverse pattern with 2 percent in the high category, 28
percent in the medium category, and 70 percent in the low (Table 19).

119

Archaeological Sensitivity

WU LW Sites (%)

WU Random Points (%)

High

63.9

2

Medium

33.3

28

Low

2.8

70

Table 19: West Unit Late Woodland Archaeological Sensitivity.
The results are different when the scores are applied to the entire eastern UP set of LW
sites (Appendix P). In this case, 37 percent of the LW sites are identified in high sensitivity
locales, 57 percent in medium sensitivity locales, and the remaining 6 percent in low sensitivity
areas. Despite the flip in the relative proportion of LW sites in high and medium sensitivity
locales, 94 percent of the LW sites can be classified in high or medium sensitivity locales. None
of the random eastern UP points fell within the high category, 21 percent are in the medium
category, and 79 percent are in low sensitivity areas (Table 20).
Archaeological Sensitivity

EUP LW Sites (%)

EUP Random Points (%)

High

37

0

Medium

57

21

Low

6

79

Table 20: Eastern UP Late Woodland Archaeological Sensitivity.
Summary statistics highlight the differences between the LW site scores and the random
point scores (Table 21). Both the WU and eastern UP LW site sets have significantly higher
mean and median scores than the random points. The mean and median scores of the WU and
eastern UP sites supports the differences observed in the previous paragraph.
Chi-square tests comparing the LW site predictive scores in the WU and the eastern UP
with random point predictive scores show statistically significant differences in their
distributions (Table 22). However, a comparison of the WU and eastern UP sites also

120

Summary Statistics

WU LW Sites

EUP LW Sites

WU Random Pts.

EUP Random Pts.

n=

36

81

50

80

125.8

112.1

37.7

28.6

Median

140

108

19

19

St. Dev.

30.9

32.6

35.1

26.3

Minimum

9

9

3

1

Maximum

155

155

140

108

Mean

Table 21: Summary Statistics Comparing Archaeological Predictive Scores.

Chi-Square Tests
WU LW sites v.
WU Random
points
EUP LW sites v.
EUP Random
points
WU LW sites v.
EUP LW sites
WU Random
points v. EUP
Random points
WU Interior sites v.
EUP Interior sites
WU Coastal sites v.
EUP Coastal sites

Observed χ²
Statistic

Critical Value @
0.1

P Value

Statistically
Significant

48.978

4.605

0

Yes

92.817

4.605

0

Yes

7.508

4.605

0.023

Yes

2.448

4.605

0.294

No

0.276

2.706

0.599

No

5.220

4.605

0.074

Yes

Table 22: Summary of Chi-Square Tests.
demonstrates a statistically significant difference in the distribution between these sets of sites.
Conversely, there is not a statistically significant difference in the distribution of the WU and
eastern UP random points. The site data indicates that there is a difference in the locations
selected for sites in the LW between the WU and the eastern UP as a whole. The random point
data suggests that the range of potential locations is similar in the WU and the eastern UP.
We already have discussed that there is a higher proportion of interior known LW sites
on the WU and a higher proportion of coastal known sites for the entire eastern UP. All of the
sites classified within low sensitivity areas are in coastal settings (there are no known LW sites in

121

low sensitivity locales in the interior). Interior sites have a higher mean predictive score (128.6
[140 median]) than coastal sites (104.8 [105 median]) (Table 23). The higher proportion of
interior sites, therefore, may be skewing the overall predictive score in the WU. When the score
of interior setting sites are compared, there is not a statistically significant difference in their
distribution (Table 22), but there is a statistically significant difference between the sets of
coastal sites. One interpretation may be that access to spawning beds at coastal sites may
outweigh other environmental factors, such as terrestrial habitat classification. Martin’s (1985)
study of Woodland coastal settlement found that site location was most closely related to fish
habitat. Alternately, and potentially related to this interpretation, the WU may simply have more
optimal interior setting habitat for LW sites than the rest of the eastern UP.
Summary
Statistics

Interior Sites

Coastal Sites

Lake Huron
Sites

Lake Michigan
Sites

Lake Superior
Sites

25

56

18

16

22

128.6

104.8

89.6

108.3

114.7

Median

140

105

105

108

108

St. Dev.

22.1

33.9

30.6

37.3

30.8

Minimum

92

9

9

24

9

Maximum

155

155

108

155

140

n=
Mean

Table 23: Summary Statistics by Site Setting.
Another observation can be gleaned through a comparison of coastal sites by Great Lake.
Summary statistics show that the sites on Lake Huron have a lower mean and median predictive
score than coastal sites on Lakes Michigan or Superior (Table 23). Further, no sites on Lake
Huron have high predictive scores and Lake Huron has the highest proportion of low scoring
sites (16.7 percent) (Figure 31). Habitat seems to be the key variable in this discussion. No
coastal site on Lake Huron is situated in mixed pine habitats. Mixed pine is the best represented

122

Figure 31: Archaeological Sensitivity of Coastal Late Woodland Site Locales by Great
Lake.
for interior sites (56 percent) and for coastal sites on Lake Superior (45.5 percent) and Lake
Michigan (50 percent). Northern hardwood habitat is the best represented (44.4 percent) for
Lake Huron sites.

4.6.1 Application of the Model
The next step in the process was to apply the model to a series of sample areas in the
HNF to gain a better understanding of the models functionality. It was determined that a 100 ×
100 m (1 ha) quadrat would serve as the basic unit of analysis in the model. The 100 × 100 m
quadrat was used because it is sufficiently fine grained for useful analysis, yet coarse enough to
efficiently calculate in the GIS. The sample areas included the random generation of ten one

123

kilometer squares on both the East and West Units, subdivided into one hectare squares, as well
as the subdivision of the 2011 and 2012 HNF cultural resource survey parcels into one hectare
squares.
The LW site predictive model was applied to ten randomly selected 1 km × 1 km
quadrats on the WU of the HNF to assess the relative archaeological sensitivity of the WU
(Figure 32). Each of the 1 km² quadrats was subdivided into one hundred 100 × 100 m squares.
The 100 × 100 m square quadrats were coded in the same manner as the sites and random points.
However, their larger size than the site points or random points created the potential that the
square could include more than one habitat or have portions at more than one distance to a
source of water. The least distance from water was used to generate the score for the square. So
if a square straddled 120 m, the score would reflect less that 120 m. Each 100 × 100 m square
was coded for the presence or absence of the eight habitat categories as well as for wild rice.
Each habitat present was assigned a score and the results were added together for the habitat
score. Thus, 100 × 100 m squares with greater habitat diversity (greater number of habitats)
have a higher habitat score.
One thousand 100 × 100 m quadrats were generated within the ten 1 km² quadrats in the
WU. A review of the summary statistics for the 100 × 100 m quadrats shows a mean predictive
score of 60.3 (39 median) with a high score of 176 and a low of 11 (Table 24). Nearly 10
percent of the 100 × 100 m quadrats were classified as high sensitivity areas for LW sites, while
45.8 percent were classified as low sensitivity (Table 25).
The model was also applied to eight areas surveyed in the WU in 2011 (Dunham and
Jeakle 2012). This is not a random sample, rather it is a biased sample derived from
archaeological surveys conducted in advance of planned HNF activities (timber sales, wildlife

124

Figure 32: Location of the Randomly Selected West Unit Quadrats.

WU Random
Quadrats

WU 2011 Survey
Areas

EU Random
Quadrats

EU 2012 Survey
Areas

n=

1000

1983

1000

2488

Mean

60.3

64.6

30.5

51.0

Median

39

51

19

51

St. Dev.

45.0

43.9

29.4

32.9

Minimum

11

1

3

1

Maximum

176

191

158

173

Summary Statistics

Table 24: Summary Statistics Comparing Test Areas.

125

Archaeological
Sensitivity

WU Random
Quadrats (%)

WU 2011 Survey
Areas (%)

EU Random
Quadrats (%)

EU 2012 Survey
Areas (%)

High

9.8

14.7

0.4

7.1

Medium

44.4

28.1

12.3

10.8

Low

45.8

57.2

87.3

82.1

Table 25: Comparison of Late Woodland Archaeological Sensitivity.
management, road improvement, etc.) (Figure 33). The survey areas were not equal in size and
combine to include 1983 one hectare quadrats. These have a mean predictive score of 64.6 (51
median) with a high score of 191 and a low of one (Table 24). Two hundred and ninety two
(14.7 percent) are classified as high sensitivity, 557 (28.1 percent) as medium sensitivity, and
1134 as low sensitivity (57.2 percent).
When the 2011 survey data set is combined with the random WU 1 km² quadrat data, a
data set of 2983 100 × 100 m quadrats was generated. This set includes 390 high sensitivity
quadrats (13.1 percent), 1001 medium sensitivity quadrats (33.5 percent), and 1592 low
sensitivity quadrats (53.4 percent) (Table 26).
Sensitivity Score

WU Test (%)

EU Test (%)

High

13.1

5.1

Medium

33.5

11.2

Low

53.4

83.6

Table 26: Comparison of Late Woodland Archaeological Sensitivity.
Two LW sites were identified in the 100 × 100 m quadrats examined in this study. Site
20AR310 (FS 09-10-03-728) is situated in a high sensitivity locale (score 173) in W10 and has
been archaeologically tested (Figure 34) (Dunham 2013; Franzen 1998). Site 20ST283 (FS 0910-02-576) is located in 2011 Survey Area 2 in a medium sensitivity setting (predictive score

126

Figure 33: Location of 2011 Survey Areas.

127

Figure 34: Quadrat W10.

128

105) (Figure 35). Site 20ST283 is known based on archaeological survey data, but was only
discovered to be a LW site during this phase of the analysis after the LW archaeological site data
set statistical analyses (Dunham and Jeakle 2012).

Figure 35: 2011 Survey Area 2.
4.6.2 East Unit Test
Ten 1 km × 1 km quadrats were also randomly selected on the EU of the HNF to
characterize the relative archaeological sensitivity of the EU. As with the WU sample, each of
these was subdivided into 100 m × 100 m squares (a total of 1000 quadrats) (Figure 36). These
were coded with a predictive score like the WU quadrats. One thousand 100 × 100 m quadrats
were generated within the ten 1 km² quadrats in the EU. A review of the descriptive statistics for

129

the EU 100 × 100 m quadrats shows a mean predictive score of 30.5 (19 median) with a high
score of 158 and a low of 3 (Table 20). Four of the 100 × 100 m EU quadrats were classified as
high sensitivity areas for LW sites (0.4 percent), 123 (12.3 percent) were medium sensitivity, and
873 (87.3 percent) were classified as low sensitivity.

Figure 36: Location of Randomly Selected East Unit Quadrats.
Seven survey areas examined in 2012 (Dunham 2013), that include 2488 one hectare
quadrats, were also explored with the model (Figure 37). The survey areas were not equal in size
and were selected based on planned HNF activity as opposed to the randomly selected 1 km²
quadrats. These have a mean predictive score of 51.0 (51 median) with a high score of 173 and a
low of one (Table 24). One hundred and seventy six (7.1 percent) are classified as high
sensitivity, 269 (10.8 percent) as medium sensitivity, and 2043 as low sensitivity (82.1 percent).

130

Figure 37: 2012 Survey Area Locations.

131

Three thousand four hundred and eighty eight EU quadrats were examined in this study and
include 180 high sensitivity quadrats (5.1 percent), 392 medium sensitivity quadrats (11.2
percent), and 2916 low sensitivity quadrats (83.6 percent) (Table 22). None of these quadrats
included a LW site.

4.6.3 Discussion
From a practical standpoint, The WU random test quadrats provide some useful
illustration of some of the potential weaknesses of the model. For example, Bishop Lake Creek
is not coded as a stream within the GIS on W7, so the areas along it are not coded as near a body
of water and the areas are depicted as low sensitivity (Figure 38). Similarly, there are medium
sensitivity areas in W6 which are in Little Lake 16 and should not have sensitivity for LW sites
(Figure 38). These quadrats are within 120 m of the lakeshore, but are also under water. Each of
these examples highlights the importance of field truthing locations and traditional prefield
background review of survey areas as opposed blind reliance on the model.
The application of the model also illustrates some potentially significant trends in the
WU and the EU as well as between these two parts of the HNF. The 2011 and 2012 survey
samples include higher proportions of high and medium sensitivity areas than their random test
quadrat counterparts, and a lower proportion of low sensitivity areas (Table 25). The best
explanation of this is that the 2011 and 2012 survey samples were drawn from the cultural
resource surveys carried out by the HNF in advance of their planned activities. One of the most
important activities on the HNF is timber management, including the harvesting of timber. The
management and harvest of pine for pulpwood and lumber is the most significant activity in
timber management. Both the 2011 and 2012 survey areas included large areas of pine

132

Figure 38: Quadrats W6 and W7.
plantation (Dunham and Jeakle 2012; Dunham 2013). While pine plantations are not always in
mixed pine habitats, that habitat is optimal for red and white pine, and it is possible that survey
areas selected by the FS are skewed towards habitats more conducive to pine management.
Higher predictive scores may be seen in FS generated survey areas as a result.
When the WU and EU samples are compared, there are significantly higher proportions
of high and medium sensitivity quadrats identified in the WU and a higher proportion of low
sensitivity quadrats in the EU (Table 26). The mean predictive scores are also higher for the WU
than the EU (Table 24). These data suggest that high and medium sensitivity areas for LW sites
are better represented on the WU than the EU and this is likely the critical factor in the higher
number of sites on the WU.

133

4.7

Application of the Model to the HNF
The final step in the development of the LW site predictive model for the HNF was to

apply the model to the entire Forest. The entire HNF was subdivided into 100 × 100 m squares
(1 hectare quadrats). The quadrats were scored in the same manner as presented in the preceding
section (Figures 39 and 40). When the model is applied to the entire WU and EU areas 530,593
one hectare quadrats are observed.
Based on the predictive scores, some of the areas included are likely open water (inland
lakes, as per the example of Little Lake 16 above, and portions of the Great Lakes along the
coasts). Overall predictive scores of 1, 9, or 90 were deleted as likely water, providing they did
not include terrestrial habitats. When the areas that are open water are removed, the number of
quadrats is reduced to 523,400 including 318,557 quadrats on the WU and 204,843 quadrats on
the EU.
The WU has proportionally greater archaeological sensitivity than the EU (Table 27).
The WU, overall, includes proportionally more high and medium LW sensitivity areas (4.0 and
26.2 percent, respectively), as well as a higher mean score (49.4 mean, 27 median), than the EU
(2.6 and 23.6 percent, 45.8 mean and 24 median).
An examination of the WU quadrats that included LW sites parallels the distribution of
the LW site points. About two-thirds (67.6 percent) of the WU LW sites in high sensitivity
areas, 29.7 in medium sensitivity areas, and 2.7 in low sensitivity areas. In the EU, 31.3 percent
are in high sensitivity area, 68.7 in medium sensitivity areas, and no sites in low sensitivity areas.
For the entire HNF, 56.6 percent of the LW sites are in high sensitivity quadrats, 41.5 percent in
medium sensitivity quadrats, and 1.9 percent in low sensitivity quadrats.

134

Figure 39: West Unit Archaeological Sensitivity.

135

Figure 40: East Unit Archaeological Sensitivity.

136

There is also a higher density of known LW sites in the WU compared to the EU. Thirty
seven LW sites are present in the WU and 16 LW sites are present in the EU. When land area is
considered, LW sites occur at a 1.5 times greater rate on the WU than the EU. When we
compare the probability based on the relative sensitivity of a quadrat, one is about twice as likely
to encounter a LW site in a high sensitivity area on the WU and 1.7 times more likely in a
medium sensitivity area on the EU than corresponding areas on the WU. The probability of
finding a LW site on the WU of the HNF are less than a quarter percent in a given high
sensitivity quadrat or only about 1 site per 504.9 high sensitivity quadrats/hectares (Table 28). If
these figures are representative of the LW site sensitivity on the HNF and the eastern UP, LW
sites are a rare commodity.

Sensitivity Score

West Unit (%)

East Unit (%)

HNF (%)

High

4.0

2.6

3.4

Medium

26.2

23.6

25.2

Low

69.8

72.8

71.4

Table 27: Archaeological Sensitivity by HNF Unit and Overall.
WU Sites per
Quadrat

WU Quadrats per
Site

EU Sites per
Quadrat

EU Quadrats per
Site

High

0.00197

504.9

0.00093

1069.6

Medium

0.00013

7588.7

0.00023

4389.1

-

222,457

-

-

Sensitivity Score

Low

Table 28: Likelihood of Encountering a Late Woodland Site.
The difference in the distribution of high, medium, and low sensitivity areas in the EU
and WU is statistically significant (Observed χ² = 1284.686, Critical Value @ 0.1 = 4.605, P
Value = 0). This suggests that the WU includes a greater proportion of favorable locations for

137

LW site placement and this is borne out by the higher proportion of LW sites on the WU. A
puzzling factor in this discussion is the higher density of LW sites in medium sensitivity areas on
the EU.

4.8

Discussion
Phase I surveys on the HNF have discovered LW sites, but most of the information

involves location and the presence of LW diagnostic artifacts. The goal of the current chapter
was to develop a predictive model for LW sites in the HNF that would allow the determination
of spatial patterns of archaeological sites at a regional scale and to develop a ranking system to
assess those patterns. It was concluded that a correlation does exist between specific
environmental variables (distance to water, habitat, and proximity to wild rice locales).
The application of the model successfully plotted the location of 67.6 percent of the
known West Unit LW sites in high sensitivity areas with 29.7 in medium sensitivity locales and
2.7 in low sensitivity areas. Although the model was not as successful for predicting known sites
on the EU, when the model is applied to the entire HNF it accounted for 56.6 percent of known
LW sites in high sensitivity areas, 41.5 percent in medium sensitivity locales, and 1.9 percent in
low probability areas. For the eastern UP, 37 percent of the LW sites are identified in high
sensitivity locales, 57 percent in medium sensitivity locales, and the remaining 6 percent in low
sensitivity areas. Although the model was not an ideal fit for the entire eastern UP, it does
demonstrate that 37 percent of all LW sites in the eastern UP are situated in areas that make up
only 3.4 percent of the estimated UP land base (if we can use HNF lands as a proxy).

138

The observation that the WU has a higher proportion of LW sites than the EU as well as a
higher proportion of high and medium sensitivity quadrats, was not entirely unexpected based on
more generalized observations of prehistoric site distributions on the HNF, but had not been
systematically quantified prior to this study (see Dunham and Branstner 1998; Franzen 1983).
However, the higher incidence of LW sites in medium sensitivity areas on the EU, and the
eastern UP in general, raises an interesting series of questions. This raises the possibility that
there are cultural factors or currently unrecognized environmental factors which effect the
placement of sites beyond the natural environmental variables explored as part of this site
location model. These differences could also reflect Forest Service management strategies. The
relative proportion and relative quality of archaeological survey has been comparable on the East
and West Units. Despite this, only 6 EU LW sites have had test excavations carried out on them
(50 percent of the LW sites on the EU), whereas 27 WU LW sites have had test excavation (75
percent o the LW sites on the WU). It would be worthwhile to target some high sensitivity areas
on the EU and medium sensitivity locales on the WU for additional archaeological survey. Also,
known prehistoric sites on the EU that have not produced diagnostic artifacts could be further
investigated to obtain additional archaeological evidence as to the age of the sites.
The model also provides an assessment tool for more fully exploring the setting of known
sites. A modified form of catchment analyses can be easily derived from the scored 100 x 100 m
squares. Using the LW site point as the center of the catchment, scores for each of the squares
included within the catchment can be generated. In addition to the predictive scores, it is
possible to observe the proportion of each habitat type within the catchment and provide a
measure of habitat diversity. These data, combined with the DUI scores from the previous

139

chapter (Chapter 3) will be explored in the following chapter (Chapter 5) in an attempt to
understand the dynamics of the LW settlement and subsistence system(s) in the eastern UP.

140

5.0

DIVERSITY USE INDEX AND LATE WOODLAND SITE PREDICTIVE
MODEL
Late Woodland site assemblages were summarized in Chapter 3 and the environmental

settings of those sites were discussed in Chapter 4. In this chapter these analyses are combined
to explore LW settlement dynamics at three scales. The first is a comparison of the diversity use
index (DUI) and LW site predictive model (LWPM) at the site level. The second and third carry
out a sort of catchment analysis using the DUI and LWPM as well as a site sensitivity and habitat
diversity (SS&HD) measure developed in this chapter. The catchments used in this analysis are
150 m radius and 1500 m radius. The goal of this chapter is to explore how the settings and
assemblages of LW sites may facilitate our understanding of the LW settlement patterns.

5.1

Combining the Diversity Use Index and Late Woodland Site Predictive Model
The analyses carried out in the previous chapters created two scales for assessing LW

archaeological sites. The DUI developed three ranked classes of LW sites reflecting the range of
diversity in the artifact assemblages. Sites with extended diversity included the greatest diversity
of artifact types in their assemblages. The sites with extended diversity most likely represent
residential sites which are expected to have a longer duration of occupancy as well as larger
populations that include mixed gender and age groups that are involved with a wider number of
activities (c.f., Binford 1980; 1983). These sites are most likely the seasonal aggregation sites
described by Cleland (1982) where spring and/or fall fishing took place (see also S. Martin
1985).
Limited diversity sites exhibit the fewest activities and probably represent logistical
camps used by smaller groups for limited periods of time (c.f., Binford 1980, 1983). These sites
likely would have been used for specific resource procurement activities or for specialized
141

purposes by a potential age and/or gender exclusive group. Intermediate diversity sites could
also represent residential sites, albeit with a smaller population or shorter occupation, or
logistical camps.
The LWPM created a three tiered ranking to identify settings or locations of greater
archaeological sensitivity for LW sites in the Hiawatha National Forest. This model concluded
that a correlation exists between specific environmental variables and site locations (distance to
water, habitat, and proximity to wild rice locales). Although the model was not an ideal fit for
the eastern UP as a whole, it does demonstrate that 37 percent of all LW sites are situated in
habitats that make up only about 3.4 percent of the eastern UP land base. The DUI and LWPM
scales measure two different aspects of LW settlement, but can be used in concert to better
understand LW settlement dynamics in the eastern UP.
Table 29 shows the raw counts per DUI and LWPM rank classifications. Not all the LW
sites examined have both DUI and LWPM scores and only where a site has both was it counted
in the table (n=76) (see also Appendix Q). LW sites in medium sensitivity areas and limited
diversity sites are the best represented categories. When the proportion (percentages) of sites are
compared by category it is evident that high sensitivity areas are best represented among limited

High LWPM

Medium LWPM

Low LWPM

Extended DUI

2

6

1

Intermediate DUI

6

13

0

Limited DUI

20

25

3

Table 29: Late Woodland Sites per Diversity Use Index (DUI) and Predictive Model
(LWPM)

142

diversity sites and least well represented among extended diversity sites (Figure 41). Medium
sensitivity areas are the best represented for all DUI categories. Finally, low sensitivity areas are
only found with extended diversity and limited diversity sites.
When the LW sites are examined by coastal or interior settings, these trends become
more evident (Figure 42). First, there are no extended diversity sites on the interior, which was
discussed in Chapter 3. Second, there are no interior sites in low sensitivity areas (see Chapter 4).
Finally, medium sensitivity locales are best represented for all DUI categories except interior
limited diversity sites, where high sensitivity areas are best represented.

80
70
60
50
High
40

Medium

30

Low

20
10
0
Extended

Intermediate

Limited

Figure 41: Comparison of Diversity Use Index and Late Woodland Site Predictive Model
by Category (percent).

Limited diversity sites make up the largest proportion of the LW site sample (63.2
percent). With the assumption that limited diversity sites are most likely logistical camps, then it
stands to reason that interior logistical camps are targeting locations with the best access to
resources. All high sensitivity locales are proximate to water (within 120 m) and are situated in

143

mixed pine habitats. In general, these habitats would have access to a wide range of resources
that were attractive to LW peoples as well as the animals they hunted (see Chapter 4.1.3). Fish
live and spawn in water, and some of the best represented mammal species, such as beaver and
moose, require direct access to water. High sensitivity locales seem to best exemplify the setting
of a “site for all seasons” described by S. Martin (1999) where multiple resources could be
hunted, fished, or gathered at multiple times of the year.

80
70
60
50
High

40

Medium

30

Low
20
10
0
Extended

Intermediate

Limited

Extended

Coastal

Intermediate

Limited

Interior

Figure 42: Comparison of Diversity Use Index and Late Woodland Site Predictive Model
by Category and Site Setting (percent).

The relationship between the higher incidence of low sensitivity areas and coastal sites
was explored in Chapter 4, but is worth revisiting. It seems reasonable that access to spawning
beds at coastal sites may be more important than access to terrestrial habitats. Coastal LW site
location is thought to be closely related to fish habitat (S. Martin 1985). This may also be
supported by comparing interior sites situated on lakes and rivers. The interior sites are nearly
equally distributed between lakes and streams (12 and 11, respectively). They are also similarly
144

distributed amongst intermediate and limited diversity sites (25 percent intermediate/75 percent
limited and 27.3 percent intermediate/72.7 percent limited). The difference is that 75 percent of
interior lake LW sites are found in high sensitivity areas, whereas 63.6 percent of LW sites on
rivers are found in medium sensitivity locales (Figure 43). In other words, interior lake sites are
most often situated in mixed-pine habitats in close proximity to water and interior river sites are
typically situated in other habitats.

8
7
6
5
High

4

Medium

3

Low

2
1
0
Intermediate

Limited

Intermediate

Lake

Limited
River

Figure 43: Comparison of Diversity Use Index and Late Woodland Site Predictive Model
by Category and Interior Site Setting (count).

Franzen (1987) identified some general trends concerning prehistoric interior setting sites
in the eastern UP, although his study is not specific to the LW period. He hypothesized that
interior sites on rivers were likely used for spring fishing. The presence of a river, as the source
of spring-spawning fish, was the primary factor in site placement (Franzen 1986; 1987). His
study also found that most prehistoric sites on interior lakes were associated with pine-oak forest
which is directly comparable to the mixed pine habitats used in this study (Franzen 1987;

145

Dunham 2009). The forest succession pattern in the pine-oak forests is conducive to beaver,
moose, and warm season deer habitat because of the regular creation of openings, especially
through fire, and a high incidence of aspen and other habitat classification preferred by these
herbivores (Franzen 1986, 1987). The primary resource draw for the interior river sites was
spring-spawning fish, not necessarily the surrounding habitat of the locale. Interior lakes, on the
other hand, would not have been as optimal for spring fishing as river locales, but offered
habitats conducive to other potential resources. This pattern appears to be repeated in the current
study. Thus, lower sensitivity settings, both on the Great Lakes coasts and in the interior, may
reflect an aquatic orientation (access to spring and fall spawning fish). Higher predictive scores
may reflect a broader range of resource procurement, especially for interior sites and limited
diversity (logistical) sites.
Further trends can be gleaned when the relative age of the sites are compared. Based on
the set of LW sites that includes both LWPM and DUI ranks, there are 20 sites that include an
early LW component, 43 that include a late LW component, and fourteen of the sites overlap
with both early and late LW components (Appendix Q). Table 30 presents counts in each DUI
and LWPM category and Figure 44 displays these as percentages. The most striking difference
is that early LW extended diversity sites are more proportionally balanced compared to the more
proportionally dispersed late LW extended diversity sites. Related to this is a higher proportion
of high sensitivity locales for early LW extended diversity sites and a lower proportion of
medium sensitivity areas for the intermediate sensitivity sites (the same site, 20CH6, represents
the low sensitivity proportion for both early and late LW extended diversity categories). Another
important distinction between the early LW and late LW sets is that the late LW also includes
low sensitivity areas in the limited diversity category.

146

When the early and late LW data is compared by coastal and interior settings, some of
these trends become increasingly clear and new trends appear (Table 31). The first thing to note

High LWPM

Medium LWPM

Low LWPM

Extended DUI

2

2

1

Intermediate DUI

1

6

0

Limited DUI

2

6

0

Extended DUI

2

4

1

Intermediate DUI

3

8

0

Limited DUI

10

13

2

Early LW (n=20)

Late LW (n=43)

Table 30: Early and Late Late Woodland (LW) sites per Diversity Use Index (DUI) and
Site Predictive Model (LWPM).

90
80
70
60
50

High

40

Medium

30

Low

20
10
0
Extended

Intermediate

Limited

Extended

Early LW

Intermediate

Limited

Late LW

Figure 44: Comparison of Diversity Use Index and Late Woodland Site Predictive Model
by Category and Relative Age (percent).

147

is that there is only one early LW site (20DE378) in an interior setting, and that there are ten late
LW sites located on the interior. The second is an increase in the proportion of medium
sensitivity areas and a decrease in high sensitivity locales for late LW intermediate and limited
diversity coastal sites (Figure 45). The proportions of the extended diversity coastal sites remain
constant as there are no extended diversity sites on the interior. Late LW interior sites include a
higher proportion of high sensitivity areas and do not include any low sensitivity areas.

Early LW (n=20)

High LWPM

Medium LWPM

Low LWPM

Extended DUI

2

2

1

Intermediate DUI

1

5

0

Limited DUI

2

6

0

Extended DUI

0

0

0

Intermediate DUI

0

1

0

Limited DUI

0

0

0

Extended DUI

2

4

1

Intermediate DUI

1

7

0

Limited DUI

4

12

2

Extended DUI

0

0

0

Intermediate DUI

2

1

0

Limited DUI

6

1

0

Coastal

Interior

Late LW (n=43)

Coastal

Interior

Table 31: Early and Late Late Woodland (LW) sites per Diversity Use Index (DUI) and
Site Predictive Model (LWPM) and by Site Setting.

148

100
90
80
70
60
50

High

40

Medium

30

Low

20
10
0
Extended

Intermediate

Limited

Extended

ELW Coastal

Intermediate

Limited

LLW Coastal

Figure 45: Comparison of Diversity Use Index and Late Woodland Site Predictive Model
Coastal Sites by Category and Relative Age (percent).

The preceding discussion adds onto the results of Chapters 3 and 4 to illustrate a set of
trends relating to LW settlement practices in the eastern UP. The results of these comparisons
support the relationship between lower sensitivity areas and coastal sites. It also indicates that
greater use of the interior is more common in the late LW than the early LW. There are a higher
proportion of extended and intermediate diversity early LW sites and a lower proportion of
limited diversity sites compared to late LW sites (Figure 45; Table 31). This pattern may reflect
a greater reliance on residential mobility by early LW people and a greater use of logistical
mobility by late LW people. A chi-square test does not support a statistically significant
difference between the distribution of early and late LW sites, but the graph in Figure 45
suggests a greater use of limited diversity sites by late LW people. (Yates χ² = 1.812, df 2, α 0.1
= 4.605, P = 0.404). These themes will continue to be explored through the catchment
discussion that follows.

149

5.2

Late Woodland Site Predictive Model and Habitat Diversity Catchments

The predictive model for LW sites developed in Chapter 4 can also be used as an
analytical tool.

For example, catchment analyses can be easily derived from the scored 100 ×

100 m squares. “In broad outline, site catchment analysis delimits a territory or set of concentric
territories surrounding a site and assesses the resource potential contained within that area”
(Roper 1979:122). Catchment analysis has been effectively used in archaeological studies in
northern lower Michigan and the eastern UP (c.f., Buckmaster 1979; Franzen 1987; Holman
1978; Martin 1977). Holman’s (1978) study evaluated resource potential in the site catchment
areas by season and compared the results of that analysis with archaeological data recovered
from early Late Woodland sites (Mackinac Phase). Another important outcome of this research
was that the seasonal resource potential of the catchment areas along with the data derived from
the archaeological assemblages generally corroborated assumptions concerning seasonal
resource selection and patterns of mobility gleaned from the ethnohistoric record (Holman 1978;
see also Holman and Krist 2001).
Using the LW site point or Random point as the center of the catchment, scores for each
of the squares included within the catchment can be generated. Because of the “squares”, and
the fact that the site point is not typically centered in a hectare square, the total number of
squares included is not consistent for each catchment area. This can be corrected for by
calculating a mean predictive score (MPS) for the catchment.
In addition to the predictive score, it is possible to observe the proportion of each habitat
type within the catchment. This can also be constructed as a habitat diversity index by
multiplying the number of habitat types by the number of quadrats containing each habitat type.

150

Like the mean predictive score, it can then be rendered as a mean habitat diversity index (MHDI)
for the catchment.
Two sizes of catchment were examined: 150 m and 1500 m radius. These were selected
as proxies for what can be described as “site specific zones” (150 m) and “primary zones of
exploitation” (1500 m) (cf., Rogers and Rogers 1976). The site specific zone is the area where
the site is situated and most site related activity likely took place, whereas the primary zone of
exploitation refers to the area that is readily accessible to the occupants of the site and where it is
presumed that a significant amount of direct procurement took place. The catchments are limited
to sites within the HNF boundaries which were formally coded and ranked for site sensitivity
(Chapter 4). In cases where site catchments significantly overlap, they were considered the same
catchment and the MPS or MHDI generated for the combined catchment. The catchment
analysis included in this study will provide a baseline description of habitats types present
around the LW sites. It does not explicitly measure the extractive value of the habitats in general
or by season. The following analysis is an expedient tool to assess the range of habitats and
relative LW site sensitivity of the site catchments.

5.2.1 150 m Radius Catchments
Descriptive statistics show that LW site catchments have higher predictive scores and
habitat diversity as well as a greater number of habitats per catchment than Random points
(Table 32). Additionally, t-Tests support these findings and demonstrate that Random points and
Late Woodland sites are derived from different populations (Table 33).

151

Random MPS

Random MHDI

LW Site MPS

LW Site MHDI

n=

1390

97

650

47

Mean

50.9

2.9

116.1

53

Median

34

2.7

123

4.6

Maximum

173

8.3

188

18

Minimum

1

1

3

1

Table 32: Descriptive Statistics of the 150 m Radius Catchments for Random Points and
Late Woodland Sites Addressing Mean Predictive Score (MPS) and Mean Habitat
Diversity Index (MDHI).
Observed t
Statistic
Random Point
v. LW site

Coastal v.
Interior

River v. Lake

Predictive
Score
Habitat
Diversity
Predictive
Score
Habitat
Diversity
Predictive
Score
Habitat
Diversity

t Critical Value
@ 0.1 (twotailed)

P value
(two-tailed)

Statistically
Significant

33.788

1.646

0

Yes

4.041

1.671

0

Yes

-1.889

1.686

0.067

Yes

-2.969

1.692

0.005

Yes

1.882

1.812

0.089

Yes

0.743

1.740

0.467

No

Table 33: Results to t-Tests for 150 m Radius Catchments.
Not only do the site catchments include a greater diversity of habitats, the habitats
associated with the LW sites are also structurally different than those associated with Random
points. Figure 46 illustrates the differences and shows that the LW site catchments are more
likely to include mixed pine habitats, wetland/marsh habitats, and wild rice habitats than the
Random point catchments.
There are forty seven 150 m radius site catchments used in this study, including 4 clusters
of multiple sites (Appendix R). Each of the clusters is situated along the shore of Lake Superior
and three of these are on Grand Island (Figure 47). The Grand Island clusters include: 20AR338,
20AR398, and 20AR400 which are each located at the head of Murray Bay;
152

60.0
50.0
40.0
30.0
20.0

LW Sites

10.0

Random Points

0.0

Figure 46: Comparison of the habitat composition between Random Points and Late
Woodland sites in the 150 m radius catchments (percent).
20AR348, 20AR350, and 20AR353 which are located at Williams Landing; and 20AR358/386
and 20AR359 which are located at the mouth of Echo Lake Creek. The final cluster is
comprised of sites 20CH32 and 20CH433 which are located along the shore of Whitefish Bay.
The 150 m radius site catchments include a mean area of 13.8 ha (14 ha median) with a
minimum area of 9 ha (2 sites) and a maximum area of 25 ha for the 20AR338 cluster (Appendix
R; Table 34). The pooled mean of the predictive scores within the catchments is 114.8 (118.6
median) with a minimum of 31.7 (20AR173/174) and a maximum of 154.1 for the Bar Lake site
(20AR437) (Appendix R; Table 34; Figure 47). The pooled mean of the habitat diversity index
for the catchments is 5.3 (4.6 median) with a minimum score of one for seven catchments and a
maximum score of 18 at the Widewaters site (20AR245) (Appendix R; Table 34; Figure 47). It
is worth noting that the Bar Lake site and the Widewaters site are situated about 1 km apart on
the same stretch of the Indian River. If the same predictive scoring is applied to the mean score

153

of the 150 m catchments, then nine are high sensitivity (19.1 percent), 36 are medium sensitivity
(76.6 percent), and two are in low sensitivity catchments (4.2 percent).

Figure 47: Map showing the locations of the 150 m radius Late Woodland site clusters as
well as selected Late Woodland sites noted in text.
150 m Site Catchments

Coastal Catchments

Interior Catchments

MPS

MHDI

Area

MPS

MHDI

MPS

MHDI

47

47

47

26

26

21

21

Mean

114.8

5.3

13.8

108.5

3.9

122.6

7.1

Median

118.6

4.6

14

108

39

123.5

5.8

Maximum

154.1

18

9

148.4

12.3

154.1

18

Minimum

31.7

1

25

57.7

1

31.7

2.6

n=

Table 34: Descriptive Statistics for 150 m Radius Catchments Addressing Mean Predictive
Score (MPS), Mean Habitat Diversity Index (MDHI), and Catchment Area

154

When the site catchments are compared there are interesting results. Figure 48 compares
coastal and interior setting sites. In general, the interior sites have higher mean predictive scores
and greater mean habitat diversity than coastal sites. T-tests confirm the statistical significance
of these trends (Table 33). An intriguing observation is that all the sites with a mean habitat
diversity of 1.0 are all situated on the coast. These are sites where there is only a single habitat
present within the catchment or, in other words, sites with no terrestrial habitat diversity. When
one considers the Inland Shore Fishery model, the fishery is the critical resource and site
placement on the Great Lakes shoreline reflects access to the fishery and not terrestrial habitats.

20.0
18.0

Mean Habitat Diversity Index

16.0
14.0
12.0
Coastal

10.0

Interior

8.0

Mean
6.0
4.0
2.0
0.0
0.0

20.0

40.0

60.0

80.0

100.0

120.0

140.0

160.0

180.0

Mean LW Site Predictive Score

Figure 48: Scatter plot comparing coastal and interior LW sites by Mean Predictive Score
(MPS) and Mean Habitat Diversity Index (MHDI) for 150 m radius catchments.

Figure 49 shows the presence/absence in percent of seven habitats within the 150 m site
catchments for coastal and interior sites. This distribution shows the greatest diversions in jack
155

pine habitats, wetland/marsh habitats and wild rice habitats. Jack pine is best represented on
coastal sites, in fact none of the interior sites include this habitat. Wetland and wild rice habitats
are better represented on interior sites.
100
80
60
40
Coastal
20

Interior

0

Figure 49: Comparison of the habitat composition between Coastal and Interior Late
Woodland sites in the 150 m radius catchments (percent).
If we consider interior LW sites, and compare sites located on rivers or streams and sites
on inland lakes, we find that interior lake sites have higher mean predictive scores and slightly
higher mean habitat diversity (Table 35). T-tests show a statistically significant difference in the
MPS, but do not for the MHDI (Table 33). This pattern may also be related to fishing. As noted
in the last section, Franzen (1987) hypothesized that interior sites on rivers were likely used for
spring fishing, thus the presence of a river, as the source of spring-spawning fish, was the
primary factor in site placement. A secondary factor for interior river sites may include access to
winter deer yards, a hypothesis derived from the association between these sites and lowland
conifer environmental settings (Franzen 1987; see also Van Deelan 1996). He also hypothesized
that interior lake sites were more generalized use locales reflecting more diverse hunting and

156

gathering activities. The 150 m radius catchments generated through the model seem to confirm
these observations for site specific zones in the Late Woodland.

Lake Site Catchments

River Site Catchments

MPS

MHDI

MPS

MHDI

11

11

9

9

Mean

134.0

7.9

109.7

6.5

Median

132.7

10

111.6

5.6

Maximum

154.1

13.6

148.7

18

Minimum

119.2

2.6

31.7

2.9

n=

Table 35: Descriptive Statistics for 150 m Catchments Addressing Mean Predictive Score
(MPS), Mean Habitat Diversity Index (MDHI) for Interior Sites.
Interior lake catchments are more likely to include mixed pine, wetland/marsh, and wild
rice habitats, whereas river site catchments are more likely to include wetland conifer habitats
(Figure 50). Interestingly, wetland/marsh and lowland conifer habitats appear in over half of the
catchments of both river and lake site categories. The mixed pine and wild rice habitat
associated with interior lakes sites corresponds with the broad resource potential for interior lake
sites (Dunham 2008, 2009; Franzen 1986, 1987; see also S. Martin 1999). The association
between interior river catchments with lowland conifer habitat supports Franzen’s (1987)
contention that interior river sites were placed to access winter deer yards as well as springspawning fish.
The increased incidence of wetland/marsh habitats associated with interior site
catchments is also significant. While this relationship may be simply a factor of many LW sites
being proximate to sources of water, water that might be bounded by wetlands, the importance of

157

100
80
60
40
Interior Lake
20

River/Stream

0

Figure 50: Comparison of the habitat composition between Interior Lake Late Woodland
site catchments and River/Stream Late Woodland 150 m radius site catchments (percent).
wetland resources in hunter-gather and low-level food producer subsistence is well known
(Gallagher and Arzigian 1994; Lovis et al. 2001; Nicholas 2006). This pattern was noted in the
one km² quadrat analysis in Chapter 4.3, but is more explicit as a result of the catchment. Marsh
habitats attract a wide variety of wildlife including waterfowl, muskrat, beaver, moose, and some
varieties of spring spawning fish to name a few (see Tables 33 and 34). These habitats are also
rich in certain plant resources including aquatic tuber, cranberries, and, in some cases, wild rice.

5.2.2 150 m Radius Catchments and DUI
The MPS of the catchments and the MHDI measure two different sets of variables. The
MPS reflect how the individual 100 × 100 m quadrats score within the catchment, assessing
habitat and distance to water (LWPM), whereas MHDI is a measure of the diversity of habitats
within the catchment. These two scales are weakly correlated (r = 0.47), but this reflects the
assumption that a higher predictive score is likely to result from a greater number of habitats (see
Chapter 4.6). Conversely, if all the 100 × 100 m quadrats in a given catchment were within 120
158

m of water and all included mixed pine habitat, then the MPS would be 140 (a high score based
on the LWPM). However, the MHDI of the catchment would be 1.0, signifying the presence of
only a single habitat type and no habitat diversity (a low score). As we noted previously, LW
sites tend to have a greater diversity of habitats than Random points, so it would be expected that
higher predictive scores combined with higher habitat diversity would be an attribute of LW site
catchments.
The two scales were plotted against one another on a scatter plot graph to examine the
interrelationship between LW site catchments (see Figure 48). Additionally, the mean of each
set was calculated and three categories were established (Appendix R). Site catchments that
exceeded the mean in both scales (MPS and MHDI) were considered to have increased site
sensitivity and habitat diversity (SS&HD) (12 catchments [25.5 percent]). Those catchments that
exceeded the mean in only one set or the other were classified as having moderate SS&HD (18
catchments [38.3 percent]). The remaining catchments, those that did not exceed the mean in
either scale were classified as having minimal SS&HD (17 catchments [36.2 percent]).
The SS&HD ranks can be compared to the DUI categories. The original DUI categories
from Chapter 3 were carried over for this analysis, despite a smaller set of LW sites and site
catchments. The 150 m catchment and DUI comparison follows the same trends as the DUI and
site location comparisons (Figure 51). Extended diversity sites favor catchments with moderate
and minimal SS&HD. Intermediate and limited sensitivity site locales include intermediate
SS&HD catchments as well as moderate and minimal catchments.

159

80
70
60
50

Increased

40

Moderate

30

Minimal

20
10
0
Extended

Intermediate

Limited

Figure 51: Comparison of Diversity Use Index and Site Sensitivity & Habitat Diversity by
Category in the 150 m Radius Site Catchments (percentage).

When the 150 m catchments are compared by coastal and interior settings, the patterns
once again parallel the DUI and site location comparisons (Figure 52). All the extended
diversity sites are on the coast. Catchments with minimal SS&HD are the most common in each
of the coastal DUI site categories. Increased SS&HD catchments comprise half of the interior
intermediate diversity site catchments and over 40 percent of interior limited diversity site
catchments. A comparison of interior lake and river settings reveals that sites on interior lakes
favor increased SS&HD catchments, whereas catchments on interior rivers include lower
SS&HD catchments (Figure 53).

5.2.3 1500 m Radius Catchments
As with the 150 m catchment, Random points and LW sites were compared for 1500 m
radius catchments. The 1500 m radius catchments were calculated on two occasions, one for the
predictive scores of LW sites and Random Points and again for the habitat diversity of LW sites
and Random points. The size of the catchments varied in the two analyses. The mean size for a
1500 m LW site sensitivity catchment is 531.3 ha (521.5 ha median) and for a habitat diversity
160

80
70
60
50
40

Increased

30

Moderate

20

Minimal

10
0
Extended

Intermediate

Limited

Extended

Intermediate

Coastal

Limited

Interior

Figure 52: Comparison of Diversity Use Index and Site Sensitivity & Habitat Diversity by
Category and Site Setting, 150 m Radius Site Catchments (percentage).
5
4
3
Increased
2

Moderate
Minimal

1
0
Intermediate

Limited

Intermediate

Lake

Limited
River

Figure 53: Comparison of Diversity Use Index and Site Sensitivity & Habitat Diversity by
Category and Interior Site Setting, 150 m Radius Site Catchments (count).

catchment is 624.9 ha (601.5 ha median). A t-Test comparing the distribution of the means of
these sets did not identify a statistically significant difference, so the seperate calculations were
maintained (Table 36).
Descriptive statistics show that the LW site catchments have higher predictive scores and
habitat diversity as well as a greater number of habitats per catchment (Table 37). Once again,
LW sites have higher MPS and higher MHDI than Random points. Finally, t-Tests demonstrate
161

t Critical Value
@ 0.1 (twotailed)

Observed t
Statistic

Catchment
Area

Random Point
v. LW site

Coastal v.
Interior

River v. Lake

Predictive
Model v.
Habitat
Coastal v.
Interior
Sensitivity
Score
Habitat
Diversity
Sensitivity
Score
Habitat
Diversity
Sensitivity
Score
Habitat
Diversity

P value
(two-tailed)

Statistically
Significant

-1.638

1.666

0.106

No

-6.161

1.711

0

Yes

73.408

1.645

0

Yes

2.974

1.675

0

Yes

-1.804

1.697

0.081

Yes

-1.824

1.690

0.038

Yes

1.736

1.771

0.106

No

0.708

1.796

0.493

No

Table 36: Results to t-Tests for 1500 m Radius Catchments.

Random MPS

Random MHDI

LW Site MPS

LW Site MHDI

67409

89

20188

34

44.8

6.1

72.9

7.9

Median

27

6.2

66

7.1

Maximum

176

12.1

191

20.7

Minimum

1

1

1

2.9

n=
Mean

Table 37: Descriptive Statistics of the 1500 m Radius Catchments for Random Points and
Late Woodland Sites Addressing Mean Predictive Score (MPS) and Mean Habitat
Diversity Index (MDHI).
that the predictive scores of Random point catchments and Late Woodland site catchments are
derived from different populations (Table 36).
Not only do the site catchments include a greater diversity of habitats, the habitats
associated with the LW sites are also structurally different than those associated with Random

162

points. Figure 54 illustrates the differences and shows that the LW site catchments are more
likely to include mixed pine habitats, wetland/marsh habitats, and wild rice habitats than the
60.0
50.0
40.0
30.0
20.0

LW Sites

10.0

Random Points

0.0

Figure 54: Comparison of the habitat composition between Random Points and Late
Woodland sites in the 1500 m radius catchments (percent).

Random point catchments. This parallels the results of the 150 m radius catchment test, but the
differences are less pronounced at the 1500 m scale which makes sense as the larger aggregate
area of the catchment is more likely to level itself out than the smaller 150 m catchment.
There are thirty eight 1500 m radius site catchments (Appendix S). These include ten site
clusters including (Figure 55): 20AR338, 20AR398, 20AR400, and 20AR406 at the head of
Murray Bay on Grand Island; 20AR348, 20AR350, 20AR353, and 20AR6 at Williams Point on
Grand Island as well as on Powell Point on the mainland; 20AR358/386 and 20AR359 at the
outlet of Echo Lake Creek; 20AR245 and 20AR437 on the upper Indian River; 20CH32 and
20CH433 on Whitefish Bay; 20DE93; 20DE167, 20DE294 on the lower Sturgeon River;
20ST14 and 20ST233 on Crooked Lake; 20ST109/110 and 20ST262 on Thunder Lake; 20MK58

163

and 20MK375 near the mouth of the Pine River; and 20MK159 and 20MK261 at the mouth of
the Carp River.
The pooled mean of the habitat diversity index for the 1500 m radius catchments is 7.8
(7.2 median) with a minimum score of 2.9 (20DE326) and a maximum score of 20.7 at 20DE106

Figure 55: Map showing the locations of the 1500 m radius Late Woodland site clusters as
well as selected Late Woodland sites noted in text.

(Appendix S; Table 38; Figure 55). The pooled mean of the LW site predictive scores is 71.5
(72 median) with a minimum score of 28.7 (20AR173/174) and a maximum score of 109.4
(20AR310) (Appendix S; Table 38; Figure 55). Applying the LWPM scoring to the 1500 m
catchments identifies 24 as medium sensitivity catchments (63.2 percent) and 14 as low
sensitivity (36.8 percent) with no catchments scoring as high sensitivity.

164

When these figures are compared to the 150 m radius catchments the 1500 m radius
predictive scores are lower and the 1500 m radius habitat diversity scores are higher. As noted
above, the larger aggregate area of the 1500 m radius catchment is more likely to have a lower
MPS as it encompasses a greater number of 100 ×100 m quadrats. This is because the HNF is
mostly comprised (71.4 percent) of low sensitivity areas (see Chapter 4 [Table 27]). Conversely,
the habitat diversity score is likely to be higher in the 1500 m radius catchment because the
larger area is more likely to include more habitats.
1500 m Site Catchments

Coastal Catchments

Interior Catchments

MPS

MHDI

Area

MPS

MHDI

MPS

MHDI

38

38

38

22

22

16

16

71.5

7.8

531.3

65.8

7.0

79.3

9.0

72

7.2

521.5

59.3

6.7

87.7

8.2

Maximum

109.4

20.7

1248

108.4

20.7

109.4

13.9

Minimum

28.7

2.9

139

37.1

2.9

28.7

4.9

N=
Mean
Median

Table 38: Descriptive Statistics for 1500 m Radius Catchments Addressing Mean
Predictive Score (MPS), Mean Habitat Diversity Index (MDHI), and Catchment Area.
Figure 56 compares coastal and interior setting site catchments (see also Table 38). In
general, the interior site catchments have higher mean predictive scores and greater mean habitat
diversity than coastal sites. This observation is also supported through t-Tests (Table 36). This
pattern continues to support the premise that the critical resource for the coastal sites is the
Inland Shore Fishery, whereas interior sites typically are placed in areas with a broader set of
resource potentials as measured through site sensitivity and habitat diversity. The difference in
the relative proportion of habitats is similar to that of the 150 m radius catchments, but they are
not as distinct in the 1500 m catchments (Figure 57). Mixed pine, wetland/marsh and wild rice
habitats are better represented for interior site catchments, and jack pine is more prevalent in
165

coastal catchments. Other habitats, including lowland conifer habitats, are similarly represented
in coastal and interior 1500 m catchments.
120

Mean LW Site Sensitivity Score

100

80

Coastal

60

Interior
Mean

40

20

0
0

5

10

15

20

25

Mean Habitat Diversity Index

Figure 56: Scatter plot comparing coastal and interior LW sites by Mean Predictive Score
(MPS) and Mean Habitat Diversity Index (MHDI) for 1500 m radius catchments.
Another observation concerning coastal and interior sites is the relative size of the 1500
m radius catchments. As noted above, the mean size of all the 1500 m site catchments is 531.3
ha based on the predictive score catchments. Coastal sites have a mean area of 385.5 ha and
interior sites a mean area of 731.8 ha. A t-test shows this to be statistically significant (Table
36). This difference can be explained when the setting of coastal sites is considered. A
significant part of the 1500 m catchment at coastal sites will include open waters of a Great
Lake. The site with the smallest 1500 m radius catchment is 20MK3/11 which is situated on a
point on a small island (Round Island) (Figure 55). The catchment is small because of the large

166

amount of open water surrounding the site. Open waters of the Great Lake fall outside of the
model and do not produce predictive scores. While interior sites may be located on interior

120
100
80
60
40

Coastal

20

Interior

0

Figure 57: Comparison of the habitat composition between Coastal and Interior Late
Woodland site 1500 m radius site catchments (percent).
lakes or have their catchment include interior lakes, their presence does not reduce the size of the
catchment like a Great Lake.
Interior lake site catchments have a higher MPS and higher MHDI than interior site
catchments on rivers and streams (Table 39). However, neither of these distributions are
statistically significant (Table 36). Once again, this probably reflects the larger size of the
catchment and the greater potential for increased habitat diversity as well as decreased site
sensitivity. The trends outlined above concerning interior lake and river sites may be generally
supported in the 1500 m radius catchment. Importantly, however, the MHDI increases with
catchment size demonstrating the potential importance of the primary zone of exploitation in
leveling the resource potential of a given site catchment area.

167

1500 m

Lake Site Catchments

River Site Catchments

MPS

MHDI

MPS

MHDI

9

9

7

7

87.9

9.4

68.3

8.4

88

8.5

72.4

8

Maximum

109.4

13.9

88.7

10.6

Minimum

32

4.9

28.7

6.4

N=
Mean
Median

Table 39: Descriptive Statistics for 1500 m Catchments Addressing Mean Predictive Score
(MPS), Mean Habitat Diversity Index (MDHI) for Interior Sites.
5.2.4 1500 m Radius Catchments and DUI

In this section the MPS and MHDI are combined and scaled as they were for the 150
radius catchments. Once again, the MPS and MHDI scales are weakly correlated (r = 0.4) for the
1500 m catchments. The two scales were plotted against one another on a scatter plot graph to
examine the interrelationship between LW site catchments (Figure 56). Additionally, the mean
of each set was calculated and three categories were established (Appendix S). Site catchments
that exceeded the mean in both scales (MPS and MHDI) were considered to have increased
SS&HD (11 catchments [28.8 percent]). Those catchments that exceeded the mean in only one
set or the other were classified as having moderate SS&HD (12 catchments [31.6 percent]). The
remaining catchments, those that did not exceed the mean in either scale were classified as
having minimal SS&HD (15 catchments [39.5 percent]).
There are thirty six 1500 m radius catchments that include LW sites and DUI scores
(Appendix S). The highest site DUI score was used for the catchment if there were multiple
scores in a cluster. Extended diversity sites do not appear in catchments with increased SS&HD
(Figure 58). Intermediate diversity sites are well represented in increased as well as minimal
168

sensitivity and diversity catchments. Finally, limited diversity sites are represented in all three
types of catchment and the largest proportions of limited diversity sites appearing catchments
with moderate and minimal SS&HD.
The three extended diversity sites are situated in coastal settings (Figure 59).
Intermediate diversity sites are better represented in catchments with increased SS&HD in the
interior and in minimal SS&HD on the coast. There are no intermediate diversity sites in
moderate SS&HD catchments on the coast. Limited diversity sites appear more frequently in
70
60
50
40

Increased
Moderate

30

Minimal
20
10
0
Extended

Intermediate

Limited

Figure 58: Comparison of Diversity Use Index and Site Sensitivity & Habitat Diversity by
Category in the 1500 m Radius Site Catchments (percentage).
catchments with minimal SS&HD on the coast and in moderate SS&HD catchments in the
interior. When interior LW sites on rivers and lakes are compared, we once again see a higher
proportion of intermediate and minimal diversity interior lake sites in increased SS&HD
catchments than interior river sites (Figure 60).

169

70
60
50
40

Increased

30

Moderate

20

Minimal

10
0
Extended

Intermediate

Limited

Extended

Intermediate

Coastal

Limited

Interior

Figure 59: Comparison of Diversity Use Index and Site Sensitivity & Habitat Diversity by
Category and Site Setting, 1500 m Radius Site Catchments (percentage).

4

3

Increased

2

Moderate
Minimal

1

0
Intermediate

Limited

Intermediate

Lake

Limited
River

Figure 60: Comparison of Diversity Use Index and Site Sensitivity & Habitat Diversity by
Category and Interior Site Setting, 1500 m Radius Site Catchments (count).

5.3

Discussion
The goal of this chapter was to explore how LW site settings and assemblages could aid

in our understanding of LW settlement dynamics. The analyses discussed above identified
170

patterns and trends that do provide insight into these processes. Some of the more distinct
patterns will be summarized as follows. The comparison of the DUI and LWPM ranks at the site
level found that the proportion of LW sites in high sensitivity areas increased as DUI decreased.
Further, limited diversity interior LW sites are the most likely to be situated in high sensitivity
areas. Limited diversity sites are thought to represent small, logistical camps where resources
might be collected, hunted, or gathered by a small group over a short period of time. High
sensitivity locales are situated adjacent to water and in mixed pine habitats, both of these
variables offer a range of subsistence related resources.
Extended diversity sites, the sites that produced evidence for the widest range of
activities, are only situated in coastal settings. These sites are interpreted to be larger, residential
sites that were occupied by a larger number of people for longer periods of time. Most of the
extended diversity sites are located in medium sensitivity areas, however extended diversity sites
include the largest proportion of sites in low sensitivity areas and the smallest proportion of sites
in high sensitivity areas when compared with intermediate or limited diversity sites. These
trends are also present in the catchment analyses (150 m radius and 1500 m radius) where none
of the extended diversity site catchments were classified as having increased site sensitivity and
habitat diversity.
The results of these analyses set up an interesting dichotomy, less diverse interior sites
associated with higher sensitivity areas and more diverse coastal sites located in lower sensitivity
areas with less environmental diversity. It is posited that the extended diversity coastal sites
represent the locations where LW peoples came together to harvest and process spring and/or fall
spawning fish (see Cleland 1982; see also S. Martin 1985). As such, these sites would have had
a larger population of people representing mixed age and gender groups for at least the spring

171

and fall spawning seasons. These people would have needed food and shelter while occupying
the site as well as the facilities to harvest and process the fish. Therefore, a diversity of tasks
would be performed at these sites. Although these sites have an extended level of diversity, the
location of these sites was dependant on access to these fisheries (aquatic habitats) and not
terrestrial habitats.
A review of the placement of interior sites supports this observation. There are no
extended diversity sites in the interior. Interior sites are located both on interior lakes as well as
rivers and streams. The sites on rivers are more oriented towards medium sensitivity locales at
the site level and more towards minimal and moderate areas of site sensitivity and habitat
diversity in the 150 m radius catchments. Interior LW sites on lakes are associated with high
sensitivity areas and catchments with increased site sensitivity and habitat diversity. Franzen
(1986) hypothesized that prehistoric sites on interior rivers were well placed for procuring spring
spawning fish. It seems that access to fish spawning habitat, whether a site is on the coast or the
interior, is a critical factor in site placement if the primarily use was harvesting spawning fish. In
this instance, coastal extended diversity LW sites and interior LW sites on rivers can be
hypothesized to have been used for fishing.
Another observation concerning the use of the landscape can be derived from the
examination of the relative age of the LW sites. The most obvious pattern associated with
relative age is the greater use of the interior by late LW people. Twenty three percent of the late
LW sites are located in interior settings, whereas only 5 percent (a single site) of early LW sites
are in the interior. Eighty percent of the late LW interior sites are in high sensitivity areas. In
addition, a smaller proportion of late LW coastal sites are classified as extended or intermediate
diversity sites and a higher proportion of late LW coastal sites are classified as limited diversity

172

sites when compared to the early LW (Figure 61). These trends are suggestive of a greater use of
logistical camps by late LW peoples, both on the coast and especially in the interior.
An important factor in this discussion is mobility. In Chapter 3, Binford’s (1980)
framework of hunter-gatherer settlement strategies was outlined in the discussion of site DUI.
This model characterizes foragers as residentially mobile and collectors as logistically mobile.
Residential mobility refers to the movement of an entire group from one location to another in
pursuit of resources, whereas logistical mobility entails smaller task groups leaving and returning
to a residential camp with resources (see also Binford 1983; Kelly 1992; Lovis et al. 2005;
Morgan 2009; C. Smith 2003; Whallon 2006). The increase in archaeologically visible limited
diversity sites on the coast and in the interior by late LW people is evidence of a higher reliance
on logistical mobility.

60
50
40
ELW Coast

30

LLW Coast
20
10
0
Extended

Intermediate

Limited

Figure 61: Comparison of Diversity Use Index by Relative Age of Coastal Late Woodland
Sites (percent).
Binford (1980; 1983; 2001) presents hunting-gathering as a successful subsistence
strategy with mobility serving as an effective security measure resulting from resource and

173

information procurement and as a buffering strategy to mitigate over exploitation of resources in
a given location (see also Davidson-Hunt and Berkes 2003; Whallon 2006). This way of
thinking leads into the third trend observed from the analyses presented above relating to the
leveling effect of the larger (1500 m radius) catchments. For example, the larger area of the
1500 m radius catchment is more likely to include more habitats. Likewise, the farther a quadrat
is from water, the lower the predictive score is likely to be. The habitat diversity should increase
and the relative site sensitivity should drop as a catchment expands.
This pattern has the potential to mitigate the environmental disparity between sites in
different settings. This may be best exemplified through interior lake and river sites. At the site
level and in the 150 m radius catchments, there was a statistically significant difference in the
terrestrial habitats (diversity and site sensitivity) associated with these sites. In the 1500 m
radius catchment, the statistical differences were no longer significant. In other words, interior
sites, regardless of riparian or lacustrine setting, have similar resource potential (as measured by
site sensitivity and habitat diversity). This demonstrates an acute awareness of the
environmental variables that make site specific zones (150 m radius catchments) preferable to
other places (such as a Random point), while considering the implications of the primary zone of
exploitation (1500 m radius catchment) to overall site setting.
In this chapter the analyses from Chapters 3 and 4 were combined to explore LW
settlement dynamics in the eastern UP. The resulting analysis revealed trends that directly
inform on LW settlement patterns that have spatial, temporal, and environmental components.
Sites where spring and fall fishing took place were typically in areas with lower site sensitivity
and habitat diversity than other sites. Interior sites were much more important during the late
LW than they were in the early LW. The interior sites were typically low or intermediate

174

diversity which indicates they were likely logistical sites. The increase in the importance of
logistical sites has important implications for how the settlement system operated in the late LW.
Finally, the discovery of the role of the primary zone of exploitation (1500 m radius catchment)
as a mitigative or buffering mechanism for LW sites in different environmental settings adds a
new dimension to our understanding of the selection of site locales. Additional discussion of
these topics as well as others will be presented in the following chapter.

175

6.0

CONCLUSIONS
Previous research on the LW in the eastern UP has emphasized the Great Lakes fishery,

especially the intensive harvest of spring and fall spawning fish. This study has revisited this
topic and integrated the results of more recent archaeological research as well as pilot studies on
these topics.

The results have illustrated a series of trends concerning LW settlement and

subsistence in the region that provide a fuller picture of LW settlement and subsistence
dynamics.
Eighty one LW sites have been identified in the eastern UP and these were quantified and
summarized in tabular form and a diversity use index (DUI) was created. The diversity index
established a framework in which to consider and compare the LW sites in the eastern UP. The
DUI framework supports the contention that LW sites were used differently than the previous
models indicated. Additionally, a LW archaeological site predictive model was created (LWPM)
that demonstrated a correlation between specific environmental variables (distance to water,
habitat, and proximity to wild rice locales) and the location of LW sites. The data from the DUI
and LWPM was combined and the results revealed trends that directly inform on LW settlement
patterns that have spatial, temporal, and environmental components.
One of the themes to arise included a distinct pattern in which sites where fishing took
place were located in areas with lower site sensitivity and habitat diversity than other sites. This
indicates that the location was selected for access to the fishery as opposed to other resources. A
second trend was that interior sites were more visible in the late LW than the early LW. In a
related point, sites on the coast typically had higher DUI and a lower LWPM score than interior
sites. Finally, it was also determined that the locations selected for sites tend to become more
similar as the catchment zone around the site expands. As the catchment expands, the relative

176

sensitivity and habitat diversity becomes more similar. These trends, as well as others, are
discussed below.

6.1

Fish, Acorns, Wild Rice, and Maize
It was posited in Chapters 1, 2, and 4 that distribution of LW archaeological sites in the

eastern UP is patterned by the decisions of LW people in relation to environmental factors they
considered culturally and economically important, specifically the location of subsistence
resources. We know that access to the fisheries requires access to water, and that the fall fishery
is only accessible on the shores of the Great Lakes. Other resources examined in Chapter 4,
acorns, maize, and wild rice also have spatial, environmental, and temporal (seasonal and interannual) constraints.
Fall spawning fish can only be accessed from deep water locales on the Great Lakes (the
rapids at Sault Ste. Marie being a notable exception). Fall spawning begins as early as October,
but typically peaks between mid-November and late December (Cleland 1982). Wild rice can
occur in slow moving rivers or sheltered locations near Great Lakes shorelines, but is most
commonly found in interior lakes with stable water levels (Aiken et al. 1988; Vennum 1988).
Wild rice typically ripens in late August or early September. Maize, in reference to the eastern
UP, is most likely to succeed in areas near the Great Lakes because of the extended growing
season in the lake effect zone (O’Shea 2003; Yarnel 1964). Maize begins to ripen in August, but
can be stored on the stalk for a period of time before harvesting. Oak occurs in a variety of
habitats in the eastern UP, but are best represented in mixed pine habitats (Dunham 2009).
Acorns begin to ripen in late summer, depending on variety, and continue to do so through
October. Based on this information, nearly any of these resources may be accessed from locales

177

on the Great Lakes coast. Wild rice and acorns could alternately be collected from interior
locales. These constraints may facilitate a better understanding of settlement and subsistence
dynamics in the region.

6.1.1 Fall Fishery
The trends outlined in Chapter 5.3, as well as those documented in Chapters 3 and 4,
directly address the basic goals of this dissertation. The inverse relationship between site
diversity and site sensitivity seems to support the premise that extended diversity coastal sites
were places where fall and/or spring spawning took place. A similar pattern is noted with
interior sites, where sites on rivers are most associated with moderate site sensitivity suggesting
that these locales were used for spring fishing. With the exception of the proto-historic
component of 20DE4, all the extended diversity sites include faunal evidence for spring and/or
fall spawning fish. Sites 20AR348, 20CH95, and 20MK1 also include stone net weights further
supporting the act of fishing on these sites. Two of the interior river sites, 20DE75 and
20DE188, produced spring spawning sturgeon remains.
Two of the extended diversity site components are early LW, four are late LW, and three
include both early and late LW components that cannot currently be differentiated based on
existing reporting (Appendix T). An exception to this is the faunal remains from the Scott Point
site (20MK22) that have been divided into three components (Mackinac, Bois Blanc, and
Juntunen Phases), although comparable information was not available for the lithic assemblage
to prepare DUI scores (T. Martin 1982). Thus, 11 sites/components are represented.
Spring spawning fish remains are present on each of the extended diversity components,
except for the proto-historic component at 20DE4. Fall spawning fish remains were recovered

178

from seven components and possible gill-net sinkers from another site (20CH95). The remains
of a whitefish variety, though not identified by species, were also recovered from 20MK261
offering the possibility that a many as nine components include evidence for the harvest of deep
water fall spawning fish. The Mackinac Phase components of 20MK1 and 20MK22 (early LW)
include fall spawning fish and if the whitefish remains at 20MK261 are lake whitefish, then each
of these early LW components include fall spawning fish. Sites 20MK1 and 20MK261 are both
situated on Lake Huron at the Straits of Mackinac and 20MK22 is located along the north shore
of Lake Michigan.
The four late LW extended diversity components that produced fall spawning fish
remains are associated with two sites (20MK1 and 20MK22). Site 20CH95, where net weights
were recovered, is located on Lake Superior on Whitefish Bay. Site 20AR348, located on Grand
Island, is the only other site to produce fall spawning fish remains, but the assemblage cannot be
divided into early or late LW components.
An additional 13 components from eleven sites have produced either fall or spring
spawning fish remains (Appendix T). Ten of these are in coastal settings and three in interior
settings. Twelve are intermediate diversity and one is a limited diversity site located in the
interior. Twelve of these have produced spring spawning fish, including all three interior sites.
Fall spawning fish were found on eight components, all of which are on the coast.
When we consider intermediate and limited diversity coastal components with fall
spawning fish remains, there is an early LW component, three late LW components, and four
components that include mixed or undetermined early and late LW components (Appendix T).
Site 20MK169/457, which includes early, late, and mixed LW components, is located at the
Straits of Mackinac on Mackinac Island. The other two late LW components are located on Bay

179

de Noc (20DE296) and Grand Island (20AR359). Two of the remaining three early/late
components are at the Straits of Mackinac (20MK54 and 20MK61) and the last one, 20ST1, is
located at the north end of Lake Michigan.
Only five of the sites (or 7 components) with fall or spring spawning fish remains are in
high sensitivity locales (29 percent). Two of these are on Grand Island, two are in the interior,
and the fifth is located along the north shore of Lake Michigan. Twenty one of the LW
components with spring or fall spawning fish remains are located on the Great Lakes coast and
three are in the interior. Eleven of the coastal components are extended diversity and ten are
intermediate diversity. One of the interior sites is limited diversity and the other two are
intermediate.
With the exception of 20MK22, all of the early LW components that have produced fall
spawning fish are in the Straits of Mackinac and 20MK22 is geographicall proximate to the
Straits. There is no definitive evidence for the harvest of fall spawning fish in the early LW
period outside of the Straits and northern Lake Michigan. It is worth noting that T. Martin
(1982, 1991) has observed that the faunal assemblages from 20MK22 and 20DE296 are less
oriented towards the fall fishery than the Juntunen site (20MK1) and include a greater proportion
of mammals in their assemblages. Similarly, neither of the late LW components at 20DE4
includes fall spawning species (Brose 1970). This parallels the discussion in Chapter 3.5 where
sites in the western portion of the study area, like 20DE4, 20DE296, and 20MK22, have a higher
ratio of mammals compared to fish in their assemblages. This may indicate a decreased
emphasis on the fall fishery in the northern part of Lake Michigan including Bay de Noc.
As noted previously, a recent reevaluation of the data relevant to the fall fishery found
that the increased reliance on fall spawning whitefish and lake trout began around AD 800 at the

180

Juntunen site, after AD 1100 in northern Lake Michigan basin, and as late as AD 1400 in the rest
of the Upper Great Lakes region (B. A. Smith 2004). The evidence reviewed here and in
Chapter 3.5 suggests that northern Lake Michigan (e.g., 20MK22) apparently included fall
spawning fish prior to AD 1100, but that it may not have been as important there as it was in the
Straits of Mackinac region (T. Martin 1982). The recovery of possible gill-net weights from
strata at 20CH95 dated to ca. AD 1294 – AD 1411, suggest that the deep water fall fishery was
being harvested there before AD 1400, but still supports a later time-frame for this technology
compared to the Straits of Mackinac. This data, while not conclusive, appears to support B. A.
Smith’s (2004) contention that the adoption of the gill-net and the expansion of the deep water
fall fishery took place from south to north over the course of the Late Woodland period.

6.1.2 Maize
Maize was directly recovered on three LW sites in the eastern UP (20CH6, 20MK1, and
20MK24). Prehistoric maize is typically assumed to require 140 frost free days to produce a
reliable subsistence crop, although it does mature in a shorter period (Demeritt 1991; Hart and
Lovis 2013; Yarnell 1964). Each of the sites that produced maize is located in the most
climactically mild part of the eastern UP with over 140 frost free days based on modern
temperatures (Chapters 3.5 and 4.1; Eichenlaub et al. 1990). At this point, it is unclear if the
evidence represents maize grown on-site or if the recovered kernels represent grain exchanged
into the area.
The closest direct evidence of maize planting and growing in the UP is an extensive ca.
AD 1400 to AD 1500 ridged field complex situated on the Menominee River along the
Wisconsin border in Menominee County, south of the current study area (20ME61 [Buckmaster

181

2004]). This area exceeds 140 frost free days (Albert 1995; Eichenlaub et al. 1990). Maize
cupules were recovered in flotation samples from the ridged field features support their function
(Buckmaster 2004; Mulholland 2000). Similar ridged fields in Wisconsin are generally
associated with Oneota occupations (Bruhy and Egan-Bruhy 2014; Gallagher et al. 1985; Moffat
1979; Overstreet 2009; Sasso 2001). No ridged fields have been recorded in the eastern UP.
The maize recovered in the eastern UP is not well dated in regard to relative age in the
LW. The best discussion is from the Juntunen site (20MK1), where eleven of the 15 maize
kernels (73.3 percent) that were found in strata assigned to the late LW Juntunen component, one
kernel from the late LW Bois Blanc component, and two kernels from the early LW Mackinac
component (McPherron 1967:189; Yarnell 1964). The contexts of the recovered maize from the
Cloudman site (20CH6) and 20MK24 doesn’t allow a finer assignation than Late Woodland. A
growing corpus of data is showing that a fuller scale adoption of maize in the Upper Great Lakes
region is later than previously accepted (Hart and Lovis 2013; O’Gorman 2007).
Isotopic evidence from human remains from Juntunen Phase ossuary contexts at the
Juntunen site includes δ 13C values that are suggestive of maize consumption (Brandt 1996:7071). This study cautions that the reliance on Great Lakes fish, especially walleye, as well as the
dietary use of domestic dog can falsely increase the estimation of maize use (Brandt 1996; see
also Katzenberg 1989; see also Chapter 3.5). Brandt (1996:71), however, believes that the mean
-18.0 δ 13C value at the Juntunen site likely represents maize consumption. Despite this, it is not
clear if the maize was grown on-site, if it came in through exchange, or if the tested individuals
consumed maize at another site(s). The location of the eastern UP on the periphery of viable
maize horticulture and the evidence from the Juntunen site suggest that the late LW would be the
most likely time frame for maize horticulture in the eastern UP in the LW.

182

The spatial analyses carried out in Chapter 4 have shown that only about 6.5 percent of
the WU of the HNF falls within areas with over 140 frost free days. Additionally, a comparison
of Random points and LW sites did not find a statistically significant difference in their
distributions relative to frost free days in the WU of the HNF and only 5.7 percent of LW sites
on the WU fall within areas with more than 140 frost free days. It should be noted, that despite
this figure about a third of all the LW sites in the eastern UP fall within the 140+ zone. Over half
of these sites (51.9 percent) are located at the Straits of Mackinac and the remainder along the
north shore of Lakes Michigan and Huron.
The frost free day variable is probably not the only factor in successful maize horticulture
in the eastern UP. Maize also requires a long enough period of warm temperatures (growing
degree days) for successful maturation (Demeritt 1991; Hart and Lovis 2013). Modern
temperature data can serve as a useful proxy for past temperatures as the temperatures during the
Medieval Climatic Optimum may not have been much warmer than modern temperatures,
whereas temperatures during the Little Ice Age were certainly cooler (Bernabo 1981). O’Shea
(2003) notes the extreme levels of interannual fluctuation in maize yields in northeastern lower
Michigan in the historic period as well as the high incidence of maize crop failure among the
Huron (every three to six years) who lived in a more climatically mild area in southern Ontario
(see also O’Shea 1989:63; O’Shea and McHale-Milner 2002). In the early eighteenth century,
during the Little Ice Age, Antoine-Denis Raudot wrote of the Chippewa of Sault Ste. Marie, as
well as peoples on the north shore of Lake Huron, that they gathered maize green because it
didn’t fully ripen (Kinietz 1965:322). Raudot attributed this to the “fog,” although it seems like
the lack of adequate growing degree days was more likely a factor.

183

6.1.2 Wild Rice
The Cloudman (20CH6) site is the only site to have produced direct evidence for wild
rice, a single grain, in the eastern UP (Egan-Bruhy 2007; See also Chapter 3.5). The potential
for LW wild rice use is based on the continued presence of wild rice in the region as well as a
correlation between locations where wild rice was documented in the nineteenth and twentieth
centuries and the locations of LW sites (Dunham 2008; see also Figure 27). As part of the
current study, a statistically significant relationship was found between LW sites and modern
wild rice stands for the eastern UP as a whole as well as a quadrat analysis of the WU of the
HNF (Chapter 4.1). The location of modern and historic wild rice beds was included as a
variable in the Late Woodland site model developed in Chapter 4.
Most of the LW sites that are situated proximate to wild rice beds are in the interior
(Appendix U; Figure 62). There are three exceptions: 20CH6; 20DE106; and 20MK90. The
wild rice stands associated with 20CH6 and 20MK90 are situated inland from the sites. The wild
rice bed associated with 20DE106 is known from a herbarium source that places it in the shallow
waters of Little Bay de Noc (Edman 1969). When the age of the sites can be established, most
include late LW components. The exception is 20MK90 which has an early LW age based on
ceramic typology. Site 20CH6 includes both early and late LW components.
The combination of late LW sites in interior settings is striking when considered with the trend
that interior sites become more visible in the late LW. Further, most of the sites associated with
wild rice are limited diversity. Ethnographic sources describe wild rice camps as locations
where families came together to harvest and process the grain (see Lofstrom 1987; Vennum
1988). Wild rice camps can be thought of as smaller aggregation locales than coastal fishing
sites. The rice camp would include a multi-generation extended family group. It is

184

Figure 62: Distribution of Late Woodland sites associated with wild rice habitats.
noteworthy that three of the interior sites associated with wild rice locales (33.3 percent) are
coded as intermediate diversity and that each of these exceeded the mean DUIrev score
suggesting an increased intensity of occupation (Appendix U). These sites, 20AR437,
20CH171/172, and 20ST109/110 represent the most likely wild rice camps.
Most of the LW sites associated with wild rice patches are in high sensitivity areas (75
percent). Only one site, 20CH6 is situated in a low sensitivity area. Interestingly, the two sites
located in medium sensitivity areas are both situated on interior lakes in the East Unit of the
HNF and represent the only interior sites in that part of the National Forest (20CH171/172 and
20MK334). With the exception of these two sites, all the other interior sites associated with wild

185

rice are in the Indian River drainage which is a major tributary of the Manistique River (Chapter
2.1).
The spatial observations are worth discussing. The two interior sites on the East Unit, as
noted above, are the only two interior LW sites known in that part of the eastern UP. Likewise,
they are associated with the only interior lakes in the East Unit that had documented wild rice.
The lower proportion of high sensitivity areas in the East Unit was discussed in Chapter 4. It is
possible that the increased resource draw of wild rice in these lakes, assuming it was present in
the LW period, is related to the placement of sites at these locations.
The association with the Indian River is also significant. About half the documented wild
rice beds in the eastern UP are within the Indian River basin as are about half of the known
interior LW sites (Dunham 2008). All the LW sites associated with wild rice locales in the
Indian River drainage are late LW and when they include identifiable ceramics it is
predominately Oneota-related, although site 20AR245 produced Point Sauble ware. Both Point
Sauble wares and Oneota-related ceramics are best represented in what is today Wisconsin, to
the south and west of the current study area. There is direct archaeological evidence for the use
of wild rice by Oneota and Point Sauble peoples in Wisconsin (Arzigian 2000; Moffat and
Arzigian 2000; Overstreet 1997).
There is reasonable evidence to hypothesize that wild rice was a potential resource
collected by LW peoples in the eastern UP by the late LW. It seems most likely that this
resource was utilized in the Indian River drainage, but there is evidence placing the grain in the
St. Mary’s River and northern Lake Huron as well.

186

6.1.4 Acorns
The most compelling evidence for an alternate starchy food resource in the eastern UP
relates to acorns. Acorns are the best represented botanical remain on LW sites in the eastern
UP, appearing in 46 percent of the floral assemblages, and acorns are well represented in the
ethnographic literature of the Upper Great Lakes region (see Chapter 3.5; Appendix I; see also
Dunham 2009 for a broader discussion of this topic). Acorn lipid residue has also been
recovered from LW ceramics suggesting formal processing of this resource (Skibo et al 2009).
Additionally, red oak is an important part of the successional sequences in mixed pine habitats,
which have the highest correlation with LW site locations (see Chapters 4 and 5; see also
Dunham 2009).
Mixed pine habitats are directly associated with 39.5 percent of LW sites (Figure 63;
Appendix V). They also appear in 57.4 percent of the 150 m site catchments and 78.9 percent of
the 1500 m site catchments (Figure 63; Appendix V). When the relative age of the sites is
considered, a higher proportion of late LW sites include mixed pine habitats. Mixed pine
90
80
70
60
50
40
30
20
10
0
1500 m
Catchments

150 m Catchments Site Locales (EUP) Site Locales (ELW) Site Locales (LLW)

Figure 63: Percentage of LW Sites with Mixed Pine Habitat.

187

habitats are better represented at interior sites, with the exception of early LW sites (there is only
a single early LW site in the interior and it is not in a mixed pine habitat) (Figure 64). The
largest proportion of interior late LW sites is in mixed pine habitats (80 percent).

There are no

LW sites in mixed pine habitats in the Lake Huron basin (Appendix V).
The area of the mixed pine habitats varies within the catchments. Table 40 shows the
mean area of mixed pine habitat for all the site catchments. At 150 m, there is a mean of 6.7 ha
of mixed pine habitat and a mean of 217.4 ha for the 1500 m site catchments. While red oak is
the best represented oak species in the eastern UP it is not common, comprising less than one
percent of the forests (Bourdo 1954; Price 1994). A rough estimate of the number of oak trees
per hectare of mixed pine habitat is possible (2.43 red oak trees per hectare of mixed pine habitat
[see Dunham 2009; after Price 1994]) (Table 40). This demonstrates that LW sites were well
positioned to access acorns, despite the relative scarcity of oak in the forests.
The floral assemblages that include oak come from mixed LW assemblages as well as
late LW assemblages. At the current time, it is not possible to assign an early or late LW date to
the acorn remains recovered from 20AR348, 20AR359, and 20CH6. Sites 20DE75 and
100
90
80
70
60
50

Coastal

40

Interior

30
20
10
0
1500 m
Catchment

150 m
Catchment

Site Locale (EUP) Site Locale (ELW) Site Locale (LLW)

Figure 64: Comparison of Coastal/Interior Sites by Percentage of Mixed Pine Habitat.
188

20ST109/110 are both late LW sites. Likewise, the acorn remains from the Juntunen site are all
attributed to the Juntunen Phase component (McPherron 1967:188). This evidence as well as the
evidence presented in Figures 63 and 64 suggests acorns were more commonly used in the late
LW.
Half the sites that have produced acorn remains are extended diversity sites (20AR348,
20CH6, and 20MK1). The remaining sites are intermediate diversity (20ST109/110) and limited
diversity (20AR359 and 20DE75). There are 30 LW sites located in mixed pine habitats and
only 2 (6.7 percent) are extended diversity sites (including 20AR348), whereas 70 percent of
them are in limited diversity sites (Figure 65). With limited diversity sites interpreted as
logistical camps, it makes sense that these sites would be placed to access a resource like acorns.
The higher proportion of acorn remains on extended diversity sites also makes sense with the
assumption that resources gathered at logistical camps would be brought to extended diversity
residential camps.

6.1.5 Discussion
The results summarized above, as well as in the previous chapters, demonstrate the
importance of spring and fall spawning fish in the diet of LW peoples in the eastern UP (see also
Cleland 1982; S. Martin 1985; B. A. Smith 2004). The evidence also suggests the potential for
integration of alternate, highly productive food resources such as wild rice and acorns into the
system, as well as the importance of large game. Likewise, there is little evidence to support the
premise that maize was a widespread part of the subsistence system of LW people in the eastern
UP.

189

Hectares

Estimated Red Oak

6.7

16.3

Coastal Mean

6

14.6

Interior Mean

7.8

18.9

Maximum

23

55.9

Minimum

5

12.2

Mean

217.4

528.3

Coastal Mean

87.6

212.7

Interior Mean

395.9

962.0

Maximum

798.0

1939.1

Minimum

7.0

17.0

Mean

150 m Catchment

1500 m Catchment

Table 40: Estimation of the number of red oak trees per hectare in Mixed Pine Habitats
(after Price 1994).
80
70
60
50
40
30
20
10
0
Extended

Intermediate

Limited

Figure 65: Distribution of Late Woodland sites in Mixed Pine Habitats by Diversity Index.
The known pattern of seasonal and interseasonal mobility also demonstrates that Late
Woodland people had the opportunity to access and use a wide variety of plant and animal
resources across the landscape in addition to spring and fall spawning fish. The procurement of
190

acorns and wild rice fits well with the seasonal subsistence round proposed by earlier scholars
without conflicting with the scheduling of fall fishing and provide a reliable buffer against the
potential failure of the fall fishery. Additionally, the role of interior sites and coastal logistical
sites can be integrated into the regional model and thought of as an integral part of a broader
pattern of resource use across the landscape. The combination of an interior resource such as
wild rice, a coastal resource like fall spawning fish, and a patchy resource like acorns may have
provided an opportunity to gather, process and store for winter consumption as well as a practical
hedging mechanism if one or more were not available in a given year.
A key component to this line of reasoning is the role of storage. Storage provides a
“means of extending the use-life of acquired resources from periods of relative abundance to
periods of scarcity” (Bursey 2001:180; see also Dunham 2000a; Holman and Krist 2001; Ingold
1983). According to Kelly (1995:120), “resources become more aggregated in space and more
constrained in seasonal availability from the equator to the arctic.” In the context of this
discussion - people who are mobile, who can access aggregated resources, and have the ability to
store surplus are at a nutritional advantage (Dunham 2000a; Holman and Krist 2001; see also
Binford 2001; Cunningham 2011; Morgan 2012). Cleland’s (1982) model explicitly identifies
storage as a critical component of the fall fishery, and ethnographic sources similarly relate the
importance of stored wild rice and acorns (see Dunham 2009; Vennum 1988).
Related to this discussion is the association of limited and intermediate diversity sites
with mixed pine and wild rice habitats. These sites, presumably representing logistical camps,
are the places where these resources are collected and likely where initial processing took place.
The increase in likely logistical camps in the late LW includes the expectation that there was an
increased reliance on storage of the resources obtained at these camps (Binford 1980, 2001).

191

These resources can be stored at residential camps, at logistical camps, or at places physically
distinct from these - so called caches (see Dunham 2000a; Morgan 2012). Evidence for caching
and storing in pits has been presented for the LW in northern lower Michigan (Hambacher and
Holman 1995; Holman and Krist 2001; Howey and Parker 2008).
The surface depressions, or cache pits, described by Howey and Parker (2008) and
Holman and Krist (2001) are asserted to be late LW in age, rather than established as such.
However, cache pit complexes provide a potential measure of storage activity since they are
visible and well documented in the ethnographic and ethnohistoric literature of the region (see
Dunham 2000a). Surface depressions are not well documented in the eastern UP and none can
be conclusively linked to the LW period (Dunham and Branstner 1998:171-172; Dunham
2000a). Notable surface depression complexes are physically proximate to LW sites, such as
20DE108, 20MK90, 20AR348, and 20AR437, but the age and function of these features is not
known (Anderton et al. 1995:62; Dunham and Branstner 1995:121-126, 1998:171; Dunham et al.
1994:35-36; Dunham et al. 2010:4-18-4-21).
The variables discussed above – logistical mobility, DUI, site sensitivity, acorns, wild
rice, maize, and fish – are well illustrated by the Cloudman site (20CH6) located on Drummund
Island. The Cloudman site is an extended diversity site situated in a low sensitivity locale. It is
the only site to have produced direct evidence for maize, wild rice, and acorns. This site also has
one of the highest diversities of plant an animal remains in this study (Chapter 3.5). The
Cloudman site represents an important residential site that was likely an aggregation point for
harvesting spring spawning fish. The diversity of the floral and faunal assemblages appears to
reflect the function of a residential site within a logistical system, where resources are brought to

192

the site. The following section continues the exploration of extended diversity sites and their
setting.

6.2

Extended Diversity Sites, Persistent Places, and Anthropomorphic Landscapes
The seven sites with nine extended diversity components have been identified as the most

likely candidates for the larger, residential sites that were used as seasonal aggregation locales
where spring and/or fall fishing took place. Each of these sites is also multicomponent sites with
earlier and/or later occupations and/or multiple LW occupations (Appendix W). For example, the
Native American occupation sequence begins at least 4000 years ago and continues through the
historic period for the Williams Landing locale on Grand Island (20AR348, 20AR350, and
20AR353) (Dunham and Anderton 1999; Dunham and Branstner 1995: Robinson et al. 1991;
Skibo et al. 2004). Likewise, the Juntunen site (20MK1) has produced evidence for Native
American occupation from about 2000 years ago to the seventeenth or eighteenth centuries
(McPherron 1967).
The long term occupation at these sites is one of the basic attributes of a persistent place.
A persistent place, as defined by Schlanger (1992:93) is “a place that is used repeatedly during
the long-term occupation of a region.” According to Thompson (2010), persistent places include
locations where there is a high concentration of desirable resources, include natural or cultural
features that structure reuse, and/or are created and maintained over an extended period of time
(see also Moore and Thompson 2012). Persistent places can also provide temporal continuity, an
anchor if you will, in an archaeological landscape where all the sites, and even all the LW sites,
are not contemporaneously occupied (c.f., Dewar 1986). If we assume that the extended
diversity sites are associated with fishing, their location on Great Lakes shorelines gives them

193

access to a high concentration of a desirable resource (the Inland Shore Fishery) along with their
long term occupation/regular reuse makes them well qualified as persistent places.
Further, when the social significance of coastal aggregation sites is considered (e.g.,
Cleland 1982; Holman and Lovis 2008; McHale-Milner 1991), then these become more than
simply resource procurement locales. The Middle Woodland component of the Juntunen site
also includes burials and the Juntunen Phase component includes ossuary burials which adds to
the social importance of the locale, beyond resource potential (McPherron 1967). In the late
Late Woodland and early historic period, ossuaries are associated with important integrative
rituals, such as the Feast of the Dead (Cleland 1971; Hickerson 1960; see also Holman and Lovis
2008; McHale-Milner 1991). None of the other extended diversity sites have the same level of
evidence of formal mortuary behavior as reported for the Juntunen site, although the lack of such
does not diminish these sites as persistent places.
Another aspect of persistent places is that the long term human occupation of the locale
can alter the physical environment. Considering the Juntunen site once again, the site locale is
interpreted to have been cleared and covered by meadow soils by AD 1050, followed by a period
of erosion and deflation, partially caused by human activity, and the development of a meadow
soil about AD 1300 (Wright as cited in McPherron 1967:37-38, 189). While not an extended
diversity site, 20DE296 may reflect a similar history. Site 20DE296 has a high DUIrev score,
suggesting an intensive occupation during the late LW period. The soils at 20DE296 are
identified as a likely boroll (Anderton et al. 1991:101). Borolls are a suborder of Mollisols
which are archetypal prairie/grassland soils (Anderton et al. 1991; Buol et al. 1989; Grunwald
2013). These environments are often influenced by fire and pedoturbation, including human

194

disturbance. Thus, the environmental setting of both Juntunen and 20DE296 may exhibit
evidence for human influenced or anthropomorphic landscapes.
If we consider the results of the catchment data in light of the potential of
anthropomorphic landscapes, there are some interesting possibilities. The sensitivity scores drop
as the area around a site expands. The smallest area of consideration, the one hectare quadrat
which includes the site, typically has the highest predictive score associated (Figure 66). As the
area around the site expands to 150 m radius and 1500 m radius catchments, it was observed the
proportion of the mean catchment scores in medium and low sensitivity areas increases. This is
in part related to a general leveling of the relative sensitivity brought about by expanding the
catchment. Another interpretation may reflect land use in the vicinity of the site.
This pattern is well illustrated by LW site locations on Grand Island (Figure 67). Note
that the LW sites are clustered in areas with high archaeological potential. The farther one might
go from the site, the lower the sensitivity. The site location and the site specific zone (the 150 m
90
80
70
60
50

High

40

Medium
Low

30
20
10
0
1500 m Site Catchments

150 m Site Catchements

1 Hectare Site Quadrats

Figure 66: Proportion of High, Medium and Low Sensitivity Areas by Percent.

195

radius catchment) include the greatest level of human activity. Is there a relationship between
the increased human activity and areas that are coded as high sensitivity locales?
There are numerous examples of how human activity can modify the landscape (Abrams
and Nowacki 2008; Delcourt and Delcourt 2004; Terrell et al. 2003). Small scale plant
management (Hildebrand 2003; Trussler and Johnson 2008; Raymond and DeBoer 2006),
patterns of residential mobility (Politis 1996), or certain landscape management practices (Miller
and Davidson-Hunt 2010) have the potential to create heterogeneous habitat mosaics which may
increase the potential for subsistence resources.
Mixed pine habitats, one of the critical factors in LW site location, are the most likely to
be affected by natural disturbances and also share many of the attributes of anthropomorphic
landscapes. Native Americans in the Upper Great Lakes region, and elsewhere, effected the
composition of forests through the use of fire (Abrams and Nowacki 2008; Albert and Minc
1987; Black and Abrams 2001; Dorney 1981; Dorney and Dorney 1989; Loope and Anderton
1998; Ruffner and Abrams 2002). Low intensity fires occurring at fairly frequent intervals
shaped forest composition around settlements. The areas that were burned contained higher
incidences of mast and fruit producing species that were commonly utilized as food. These
species tend to be either fire resistant or thrive in disturbed, including burned, environments. For
example, oak forest flourished at the expense of hemlock, sugar maple, and beech in some
locations as a result of the burning (Albert and Minc 1987; Dorney and Dorney 1989; Ruffner
and Abrams 2002). While many of these studies suggest forest and understory clearing for
horticulture as a primary rationale for the burning, habitat improvement for wildlife and other
resources, such as nuts and berries, are other likely candidates (Anderton 1999; Miller and
Davidson-Hunt 2010).

196

Figure 67: Location of Grand Island Sites in Relation to Site Sensitivity.
There is direct evidence for historic burning in northern Michigan by Native American
peoples. A study conducted by Albert and Minc (1984) demonstrated that modern stands of red
197

oak at Colonial Point were established as a result of Anishinaabek agricultural practices in the
1840s and 1850s. Charcoal recovered from plots within these stands was predominately beech
and sugar maple, indicating that the original forest had been northern hardwoods, and that Native
American burning to clear land for planting fostered the transition to oak.
Similarly, Loope and Anderton (1998) have demonstrated a much higher incidence of fire
in coastal pine stands in northern Michigan than interior stands in the eighteenth century through
early twentieth century.

The fire intervals in the interior stands seem to correspond with

naturally occurring fire regimes, where the coastal pattern is interpreted to reflect Native
American land use practices – possibly associated with the maintenance of berry patches near
settlements. Andrew Blackbird’s (1897:10-11) childhood recollection of Cross Village in the
1830s appears to reflect such a fire altered environment:

"My first recollection of the country of Arbor Croche, . . . there was nothing but small shrubbery here
and there in small patches, such as wild cherry trees, but most of it was grassy plain: and such an
abundance of wild strawberries, raspberries and blackberries that they fairly perfumed the air of the
whole coast with the fragrant scent of ripe fruit.”

Recent studies of Anishinaabek traditional landscape management practices in Ontario
show that fire was, and is, used for a variety of purposes (Davidson-Hunt 2003; Miller and
Davidson-Hunt 2010). Fire is used to clear undergrowth for gardens, to facilitate vegetation
growth (such as berries and other resources like birch bark) and for habitat improvement for wild
game. Importantly, fire is seen by these people “… as beings which possess agency and who
intentionally create order in landscapes” (Miller and Davidson-Hunt 2010:401).
The evidence outlined above shows that Native Americans in the Upper Great Lakes
region were actively modifying their landscape throughout the post-European contact period
(post AD 1600). Likewise, the evidence from Grand Island, as well as the Juntunen site and

198

20DE296, make a strong case for the similar practices in the Late Woodland period. This begs
the question of whether LW peoples were drawn to these environments or did they create them,
as at the Juntunen site and Colonial Point? The question cannot be answered for all sites at this
time, but a strong case for the role of anthropomorphic landscapes around extended diversity
sites, and possibly around intermediate diversity sites, can be made. Thus, the location of the site
becomes a more desirable as a resource procurement locale over time as well as becoming a
normative cultural landscape for the inhabitants. As such, it reinforces the persistent place status
of these sites.

6.3

Extended Diversity Sites and Ceramics
In Chapter 2 the potential for multiple settlement and subsistence strategies, both

spatially as well as temporally, was raised. Likewise, ceramic evidence for different traditions in
the late LW was cited as an example of the potential for multiple cultural traditions/ethnic
identities in the eastern UP (see Figure 2). The discussion of persistent places as well as
increased use of logistic mobility also contributes to this line of thinking.
The apparent increase in or shift towards logistical mobility in the late LW has
implications for both economic and social organization (Binford 1980; 2001). Likewise, the
adoption of the fall fishery also carries similar implications (Cleland 1983; McHale-Milner
1991).

Each of these perspectives is reflecting a need for an increased level of social

organization as a result of changing economic patterns. These changes are also related to
increased reliance on both practical and social storage. The relationship between extended
diversity base camps, where intensive fishing takes place, and dispersed logistical camps, where

199

and from which other resources are procured, is reminiscent of Hickerson’s (1962:48-49)
discussion of the role/function of the village in traditional Chippewa society.
Each of the extended diversity sites had multiple occupations over a prolonged period of
time. It is hypothesized that these represent the large seasonal aggregation sites used for fall and
spring fishing in the LW period. When the distribution of extended diversity sites is plotted on a
map, it is clear that the sites are well dispersed from one another, especially when compared to
the distribution of all the LW sites (see Figures 3 and 10). The distribution of identified ceramic
types is also revealing.
An illustration of this can be see when the identified ceramic types are plotted (Figures
68 and 69). In this case, the predominant early LW and late LW types from a given site are
shown. On the map depicting early LW, the majority of the sites include Mackinac Ware as the
best represented ceramic type (Appendix X). Notable exceptions are 20AR348 (Madison Ware),
20DE7 (Heins Creek Ware), and 20MK24 (Spring Creek Ware). Interestingly, Madison, Heins
Creek, and Mackinac wares were formerly subsumed under the classification of “Lake Michigan
Ware” and Spring Creek ware could also be categorized as such under the older typology, which
suggest a general similarity in these types (Baerreis and Freeman 1958; Fitting 1968a; Mason
1966; McKern 1931; McPherron 1967). This illustrates the broad similarity in ceramic types
during this period and appears to correspond with more generalized regional expression of
material culture.
The distribution of late LW ceramics is quite different, showing a wider range of ceramic
types. Part of this is temporal, in that certain wares precede or post date others, but it also
reflects different ceramic traditions. For example, Bois Blanc and Point Sauble Collared wares

200

Figure 68: Distribution of early Late Woodland Ceramics.
both fall early in the late LW sequence, but conform to the broad regional distribution of the later
wares. Point Sauble ware is from Wisconsin and Bois Blanc ware is part of the Juntunen
ceramic sequence. The three larger groupings, Juntunen wares, Sand Point wares, and Oneota
related-wares form fairly distinct geographic distributions, although there is some degree of
overlap.
An important secondary observation is that most of the extended diversity sites, such as
20CH95, 20CH6; 20MK1, and 20MK22, as well as some of the smaller sites, such as 20AR437,
include multiple late LW pottery types (Appendix E). The maps simply show the most common

201

Figure 69: Distribution of late Late Woodland Ceramics

identified varieties at each site. The multiple ceramic types seen at the extended diversity sites
imply interaction between these ceramic traditions. For example, 20CH95 includes Juntunen
wares, Sand Point wares, and Oneota-related wares in its assemblage, and 20AR437 includes
Juntunen wares and Oneota-related wares. Likewise, sites 20CH6 and 20MK22 both include a
high proportion of Juntunen wares in addition to the Iroquoian-related (20CH6) and Oneotarelated wares (20MK22) found at those sites. The point being that multiple ceramic types are
appearing on these sites and they are not simply reflective of a single variety.
The ceramic trends outlined above support the premise that more bounded tribal
territories had emerged by the late LW (McHale-Milner1991; O’Shea and Milner 2002; see also
Parkinson 2002). Likewise, the consistent appearance of multiple ceramic forms on most of the
202

extended diversity sites indicates exchange and interaction between these groups. It can be
assumed that some of the social networks that operated in the historic period may have
originated during the late LW (Holman and Lovis 2008). Finally, the distribution of late LW
sites by DUI category is very similar to the distribution of nineteenth century Ojibwa, Ottawa,
and Menominee villages and territories depicted by Cleland (1992a, 1992b) (Figure 70).
A final observation that can be gleaned from the distribution of late LW sites is the
coastal orientation of sites with Juntunen and Sand Point ceramics and the combination of coastal
and interior settings for Oneota-related sites. This suggests different settlement patterns and
different resource use. The Juntunen site was integral to the development of the Inland Shore
Fishery model as well as the subsequent reevaluation, so the coastal orientation of Juntunen
Phase sites is not unexpected (Cleland 1982; B. A. Smith 2004). The Sand Point sites appear to
parallel the coastal orientation of the Juntunen Phase sites, although the advent of fall fishing
appears to be later (Dunham and Branstner 1995; Dunham and Hambacher 2007; Robinson et al.
1991; B. A. Smith 2004).
Previous studies of the faunal remains at two of the Oneota-related coastal sites (20MK22
and 20DE296) include evidence for a subsistence strategy that was less reliant on the fall fishery
(T. Martin 1982, 1991; T. Martin et al. 1993). Likewise, no evidence for fall spawning fish was
recovered at 20DE4, another Oneota-related site (Brose 1970). The results of the current study
found that late LW sites in the western part of the study area had a higher proportion of large
game in their assemblages than those in the east, possibly supporting the same sort of trend.
Additionally, the current study identified acorns and wild rice as other resources that could be
more intensively harvested and stored and the Oneota-related interior sites are present in likely
habitats for these resources.

203

Figure 70: Distribution of late Late Woodland sites by Diversity Use Index.

Previous studies of the faunal remains at two of the Oneota-related coastal sites (20MK22
and 20DE296) include evidence for a subsistence strategy that was less reliant on the fall fishery
(T. Martin 1982, 1991; T. Martin et al. 1993). Likewise, no evidence for fall spawning fish was
recovered at 20DE4, another Oneota-related site (Brose 1970). The results of the current study
found that late LW sites in the western part of the study area had a higher proportion of large
game in their assemblages than those in the east, possibly supporting the same sort of trend.
Additionally, the current study identified acorns and wild rice as other resources that could be
more intensively harvested and stored and the Oneota-related interior sites are present in likely
habitats for these resources.
204

6.4

Concluding Thoughts
The shift towards harvesting deep-water fall fish in the Upper Great Lakes region and

greater use of the interior may reflect reorganization (sensu Resilience Theory) of the settlement
and subsistence patterns in response to environmental instability relating to more dynamic
variation in relative lake levels, especially in the Lake Michigan-Huron basin, after AD 900
which also broadly corresponds with the timing of the Medieval Climatic Optimum and
substantial coastal dune reactivation along the Lake Michigan shore (Figure 2; Lovis et al. 2012;
Lovis, Monaghan et al. 2012). The appearance of gill nets may also coincide with coastal dune
formation in the Lake Superior basin ca. AD 1400, but overall temperatures were cooling at this
time (Figure 2). These combined changes in the physical environment may have contributed to
near coastal resources becoming less predictable. Interior resources and deep water resources,
such as fall spawning beds, were potentially less affected by these trends than the near coastal
resources and, therefore, maintained their relative predictability (resilience).
The environmental instability in the near coastal zones was a potential catalyst for
cultural responses (release sensu Resilience Theory) that created greater stability of subsistence
resources through intensification and storage as well as greater inter- and intra-group social
networks as risk buffering strategies (reorganization/renewal sensu Resilience Theory) (see
Holman and Lovis 2008; O’Shea and Milner 2002; on differing interpretations of the inter- and
intra-group dynamic; see also Whallon [2006] on hunter-gatherer information networks and
Braun and Plog [1982] on the necessity of increased regional social interaction as the result of
local environmental unpredictability). Such an interpretation is supported by regional ceramic
trends which show a greater degree of similarity in the early LW and more diversity in the later

205

Late Woodland which are interpreted to reflect emerging social identities (c.f., McHale-Milner
1991; 1998).
Further, the more residentially mobile and more immediate return oriented economy of
the early LW was gradually supplanted by a more logistically oriented, delayed return economy
in the later LW (c.f., Binford 1980; 2001; O’Shea 1981). Increased logistical mobility has been
shown to reduce subsistence risk, including risk resulting from environmental uncertainty,
through increasing diet breadth and the ability to access multiple resource patches (Binford 1980;
Grove 2009; Lovis et al. 2005; Morgan 2009). The addition of acorns and wild rice from interior
patches to the LW diet demonstrates a form of resource diversification resulting from increased
logistical mobility (growth/exploitation sensu Resilience Theory). Increased logistical mobility
and intensification on fall resources may also reflect greater social integration and complexity.
Evidence for these patterns appears to manifest itself differently in time and space across
the eastern UP. For example, the time/space framework offered by B. A. Smith (2004) for the
adoption of gill net technology suggests people in the Straits of Mackinac region were using gill
nets before people along the south shore of Lake Superior. Similarly, evidence for greater social
integration is also more apparent earlier in the Straits than in other regions (McHale-Milner
1991; Drake and Dunham 2004). Both appear to be time transgressive south to north. I am
suggesting reorganization along the lines of a tactical shift involving the pragmatic use of fall
season subsistence resources, such as fall spawning fish, acorns, and wild rice that can be
harvested in surplus and stored to offset winter shortfalls. The surplus should not be thought of
solely as increasing production, but rather as a strategy to extend food availability and broaden
the subsistence base during the cold season when resources are scarce and/or more constrained
(see Wills 1992; Lovis et al. 2001; O’Shea 1981; O’Shea and Halstead 1989).

206

The persistent places also exhibit the active role of niche creation by LW huntergatherers in the eastern UP. This can be viewed as a form of conservation or resilience within
the system (sensu Resilience Theory) in that the prolonged use of these locales makes them
increasingly attractive to the people using them. The human occupation and activity fosters the
growth of economically beneficial plants such as berries and mast producing trees, has the
potential to attract wildlife, and figures into the formation of normative perceptions of the
landscape (Abrams and Nowacki 2008; Davidson-Hunt 2003; Delcourt and Delcourt 2004; B. D.
Smith 2007). The long term use of the Juntunen site and Williams Landing, as well as the the
other extended DUI sites, appear to reflect these trends and demonstrate the importance of niche
construction in hunter-gatherer society.
An important outcome of this study is a synthesis of LW archaeological research in the
eastern UP. Much of the recent archaeological research in the eastern UP has been derived from
Federal compliance projects and this synthesis makes this data, which is currently part of the socalled “grey literature” (Seymour 2010), available to a wider audience. The syntheses, along with
the analysis of settlement and subsistence patterns, is a significant contribution to hunter-gatherer
studies, particularly the archaeology of the Upper Great Lakes region in terms of moving away
from monolithic regional models. This synthesis also provides further perspective into the role
that aquatic resources, and more intensive use of them, play in the transformation of social
organization, mobility and territoriality (Binford 2001; Cleland 1982, 1992a; Lovis and Holman
1976; McHale-Milner 1991; Schalk 1977; Thompson and Turck 2009).
More specifically, the results of the study provide new insights into the settlement and
subsistence practices of the Late Woodland peoples in the eastern UP. While no one has claimed
that Late Woodland people in the Upper Great Lakes region only ate fish, spring and fall

207

spawning fish are emphasized as critical or key resources. Further, the coastal orientation of the
Late Woodland settlement and subsistence model, relating to both economic and social factors,
has been the dominant discourse in the region. The results of this study, as well as previously
completed pilot studies, identify acorns and wild rice as likely resources for use by Late
Woodland peoples (Dunham 2008; 2009). Each of these resources fits well with the seasonal
subsistence round proposed by earlier scholars without conflicting with the scheduling of fall
fishing, yet also provides a reliable buffer against the potential failure of the fall fishery. Such
resource based buffering strategies can, and do, effectively act in tandem with socially based
buffering systems. Additionally, the role of interior sites is more fully integrated into the regional
model and included as an integral part of a broader pattern of resource use across the landscape.
The hypothesis that Late Woodland people may have used a suite of subsistence resources as a
buffer against winter risk reorients us towards a more holistic view of Late Woodland land-use
and better contextualizes the broader based economy of these people.

208

APPENDICES

209

Appendix A:

Eastern UP Landscape Ecosystems

210

Eastern UP Landscape Ecosystems
The modern landscape ecosystems across the UP, and the Upper Great Lakes region,
have been classified by Albert (1995) and these can be used to present an overview of the
ecological variation in the eastern UP. The basic premise behind the landscape ecosystem
classification is to distinguish appropriately sized ecosystems that are distinct from one another
in abiotic and biotic characteristics (Albert 1995). In this scheme, the eastern UP is included
within a single ecosystem section (Albert 1995): Section VIII, Northern Lacustrine-Influenced
Upper Michigan and Wisconsin. In turn Section VIII is subdivided into three subsections:
Subsection VIII.1, Niagaran Escarpment and Lake Plain; Subsection VIII.2, Luce; and
Subsection VIII.3, Dickinson. Each subsection includes 2 to 3 sub-subsections that are present
in the current study area. The following discussion provides a context for subsequent
discussions concerning archaeological sites and environments in the eastern UP.
Subsection VIII.1, the Niagaran Escarpment and Lake Plain, forms the southern and most
eastern portion of the study area along the Lake Michigan, Lake Huron, and St. Mary’s River
littoral and adjoining inland areas (Albert 1995). This subsection is primarily underlain by
Silurian and Ordovician limestone and dolomite bedrock (Dorr and Eschmann 1984). The
Niagaran Escarpment forms a prominent feature arching across the landscape from Bay de Noc
to Drummond Island. This feature is also important as a source of chert which was used as a raw
material for chipped stone tools in the region (Luedke 1976).
A variety of glaciolacustrine landforms are present in Subsection VIII.1 including lake
plain, sandy deltaic deposits, and dune fields (Albert 1995; Lovis et al. 2012). Ground moraine is
also locally present. The soils in the lake plain are either lacustrine sands or lacustrine clays.
The sands range from excessively well drained to very poorly drained soils, whereas the clays

211

tend to be poorly drained. In some areas, especially near the Lakes Michigan and Huron
shoreline, bedrock is close to the surface and the overlying soils are quite thin. Elevations range
from 177 m to 317 m (580 ft to 1,040 ft). The average mean level of Lakes Michigan and Huron
is currently 177 m (580 ft).
The annual precipitation is 71 cm to 86 cm (28 in to 34 in) in Subsection VIII.1. The
growing season ranges from 128 days in the north to 175 days in the south, and south of the
eastern UP, especially along Lake Michigan (Eichenlaub et al. 1990). Prior to European
settlement, much of the subsection was lowland conifer swamp (Comer et al. 1995). Northern
white cedar was common in areas with limestone close to the surface; tamarack and black spruce
were common on poorly drained sandy soil; and a more diverse hardwood and conifer swamp
forest was found in areas of clay soil. Northern hardwood, especially beech maple forest, was
common on better drained soil. Red and white pine was locally common on sand dunes and well
drained sandy soils. Extensive marshes were present along the Lakes Michigan and Huron
shoreline as well as along the St. Marys River.
The Luce subsection, Subsection VIII.2, forms the northern two-thirds of the study area
and includes the zone along Lake Superior as well as the extensive wetland complex in the
interior of the eastern UP (the Seney Sand Lake Plain) (Albert 1995). This subsection is
underlain by Ordovacian dolomite and sandstone with includes an east-west trending Cambrian
Age sandstone escarpment which forms the Pictured Rocks and over which flows Tahquamenon
falls (Dorr and Eschmann 1984).
The area along Lake Superior includes lake plain, pitted outwash, and end moraine
(Albert 1995; Schaetzl et al. 2013). The inland, Seney Sand Lake Plain area includes poorly
drained outwash plain, deltas, and sand lake plain (Albert 1995; Schaetzl et al. 2013). The soils

212

in the lake plain, moraines and outwash along Lake Superior are generally sands. The poorly
drained areas associated with the Seney region are peats and poorly drained sands. Excessively
well drained sands are present on lake plains in these interior areas. The poorly drained areas
also include tracts of interconnected transverse dune ridges. These are large sand dunes that
were formed in the mid-Holocene, beginning as early as 5000 BC, and stabilized after 2800 BC
(Arbogast et al. 2002). These dunes have excessively well drained soils. Some of the poorly
drained peat bogs were once a series of lakes which have in filled with peat over the last 3000 to
4000 years (Futyma 1982).
Elevations range from 183 m to 378 m (602 ft to 1,240 ft). The average mean level of
Lake Superior is 183 m (602 ft). The annual precipitation is 81 cm to 86 cm (32 in to 34 in) in
Subsection VIII.2. The growing season ranges from 130 days along Lake Superior in the north
to less than 100 days in the interior (Eichenlaub et al. 1990). The interior portions of this
subsection had extensive areas of wetland and swamp, including both conifer and hardwoodconifer swamp prior to 1800 (Comer et al. 1995; Zhang et al. 2000). The areas along Lake
Superior had extensive areas of northern hardwood forest on uplands as well as areas of lowland
conifer swamp. The subsection also included large, excessively drained areas with pine-barrens.
Pine was also well represented on the interior transverse dunes (Arbogast et al. 2002; Rist 2008).
The Dickinson Subsection, Subsection VIII.3, forms the westernmost portion of the study
area extending from the southwestern corner to Lake Superior (Albert 1995). The bedrock
geology is split, with the northern third, near Lake Superior, underlain by Cambrian sandstone
and the southern part by Ordovician limestone and dolomite bedrock (Dorr and Eschmann 1984).
The Dickinson subsection includes two primary landforms: a broad till plain of ground moraine
in the south; and sandy ridges and sandstone outcrops in the north (Albert 1995; see also

213

Schaetzl et al. 2013). There is also a sandy lake plain area along Lake Superior. The soils on the
till plain are mainly loamy sands and sands predominate in the northern areas. There are
localized areas of poorly drained mucks and peats as well. The elevation in this subsection
ranges from 183 m to 396 m (602 ft to 1300 ft) with the greatest variation found in the northern
section. The average mean level of Lake Superior is 183 m (602 ft).
The growing season is about 130 days in the southern areas and along Lake Superior, and
less than 100 days in the interior (Eichenlaub et al. 1990). Annual precipitation is 76 cm to 86
cm (30 in to 34 in). Pre European settlement vegetation was primarily northern hardwood forest,
with sugar maple, hemlock and beech well represented (Comer et al. 1995). Pine was locally
well represented as were lowland conifer forests.

214

Appendix B:

Baseline Site Data

215

FS Number State Number
20AR013

UTM Northing UTM Easting Site Size (M²) M² Excavated Phase
5149257.67
535088.78
5200
8.0
Excavation

Component(s)
Multiple

Reference
Clark 1993; Jones 1993
Branstner et al. 2000; Dorwin et al. 1980
Clark 1993; Jones 1993
Rutter and Weir 1985
Dunham 2013; Franzen 1998; Rutter and
Weir 1990
Anderton et al. 2011; Clark 1993
Drake et al. 2009; Dunham et al. 1996;
Robinson et al. 1991; Skibo et al. 2009; Skibo
personal communication
Dunham and Branstner 1995; Robinson et al.
1991
Anderton 1993; Dunham and Anderton
1999; Dunham and Branstner 1995;
Robinson et al. 1991
Franzen 2000; Robinson et al. 1991
Dunham et al. 2010
Dunham 2000b; Robinson et al. 1991
Dunham et al. 1997; Goltz 1992
Goltz 1992
Goltz 1992
Dunham et al. 1997; Dunham and
Hambacher 2002
Dunham et al. 1997; Dunham and
Hambacher 2002
Franzen and Drake 2005
Bigony 1968; Franzen and Drake 2005
Anderton 1987 (letter)
Janzen 1968
Demers 1991
Bigony 1968; Luedke 1976
Drake and Dunham 2007
Franzen 1975
Franzen 1975
Drake and Dunham 2007

03-028/029 20AR173/174
20AR210
03-667
20AR245

5134852.35
5156752.92
5119839.77

538224.97
549130.24
527828.47

100
7000
400

3.0
6.0
7.6

Excavation
Excavation
Excavation

Single
Single
Single

03-728
-

20AR310
20AR330

5123436.44
5154937.51

525291.85
542750.35

600
600

12.2
1.0

Excavation
Survey

Multiple
Multiple

03-754

20AR338

5146491.63

526348.95 -

-

Excavation

Multiple

03-803

20AR348

5144412.40

525265.19

5625

23.8

Excavation

Multiple

03-825
03-811
03-820/913
03-821
03-929
03-931
03-937

20AR350
20AR353
20AR358/386
20AR359
20AR398
20AR400
20AR406

5144616.21
5144624.53
5149659.72
5149693.35
5146518.10
5146520.25
5146776.49

525434.34
525499.51
524214.96
524085.18
526604.31
526774.42
526998.36

14400
900
625
1325
700
200
100

4.3
11.4
17.4
9.4
0.5
0.3

Excavation
Excavation
Excavation
Excavation
Excavation
Survey
Survey

Multiple
Multiple
Single
Multiple
Single
Multiple
Single

03-974

20AR435

5142180.06

510033.61

100

1.1

Excavation

Single

5120528.54
5144639.66
5143539.76
5116978.90
5150114.90
5104191.16
5179076.56
5146178.07
5162940.31
5095124.60
5146244.92

527087.89
527632.17
525723.79
651811.22
657434.60
729894.48
653879.42
669845.61
637266.49
725302.25
670042.45

200
500
375
1500
2000
900
200
2900
100
225
800

6.5
8.0
2.0
1.0
255.0
2.0
1.0
1.8
1.0
1.0
0.5

Excavation
Excavation
Excavation
Survey
Excavation
Survey
Survey
Survey
Survey
Survey
Survey

Single
Multiple
Multiple
Single
Multiple
Single
Single
Multiple
Single
Multiple
Single

03-976
20AR437
03-832
20AR495
03-004
20AR6
20CH171
04-001
20CH2
20CH238
20CH27
04-010
20CH32
20CH41
20CH43
04-455
20CH433
Table 41: Baseline Site Data

216

Table 41: (cont'd)
FS Number State Number
20CH45
20CH46
04-417
20CH492
20CH6
20CH77
04-023
20CH86
04-012
20CH95
20DE1
01-076
20DE106
01-080
20DE108
01-292
20DE167
20DE17
02-366
20DE188
20DE19
01-312
20DE236
01-334
20DE294
01-328
20DE296
02-414
20DE326
20DE333
01-367
20DE378
20DE4
02-035

20DE43

5102189.82

535152.56

800

13.0

Excavation

Multiple

02-549
02-015
-

20DE459
20DE50
20DE7

5102532.69
5104485.42
5061233.37

538508.33
536961.58
526165.62

1200
400
200

9.2
2.5
1.0

Excavation
Excavation
Survey

Multiple
Single
Single

Reference
Franzen 1975; McHale-Milner 1998
Franzen 1975
Dunham 2000b
Branstner 1995; Cooper 1996
Fitting 1975b
Rutter and Weir 1989; Dunham 2013
Dunham and Hambacher 2007
UMMA files
HNF CRI form nd
Dunham et al. 1994
Anderton 1993: Rutter and Weir 1986
Bianchi 1974; Richner 1973
Rutter et al. 1984
Halsey personal communication
Rutter and Weir 1989; 1990
Anderton et al. 1991
Anderton et al. 1991
Anderton et al. 1991
OSA files
Dunham and Branstner 1993
Brose 1970
Dunham et al. 2010; Franzen 1979; Rutter
and Weir 1989
Dunham and Branstner 1997; Dunham et al.
2010
Franzen 1987
Fitting 1968a; Luedke 1976

01-072
01-061/62

20DE75
20DE85

5082414.68
5076146.47

523134.83
501187.27

1250
8750

23.7
2.5

Excavation
Excavation

Single
Multiple

Buckmaster 1983; Martin and Martin 1980
Rutter and Weir 1986

01-069

20DE93

5078489.37

525215.44

3600

2.0

Survey

Multiple

05-305

20MK1
20MK102
20MK159

5076598.08
5080666.99
5099439.78

687719.60
678119.46
678056.59

7432
1000
4500

441.3
11.9
4.3

Excavation
Excavation
Survey

Multiple
Multiple
Multiple

Anderton 1993; Franzen 1998; Weir 1981
McPherron 1967; Fitting 1975a; McHaleMilner 1998
Fitting 1978
Rutter and Weir 1986

UTM Northing UTM Easting
5104567.14
750448.51
5094883.36
725530.19
5148580.61
674872.06
5103877.78
757655.66
5152581.31
704844.05
5150567.00
681827.02
5149875.66
651121.91
5061947.97
526047.09
5084015.21
503802.25
5076603.47
521821.67
5079162.36
524538.74
5075813.79
535875.51
5088081.22
522882.31
5062882.70
525644.07
5057902.38
502676.80
5077402.40
525732.21
5076411.83
516806.79
5107360.53
537484.96
5046353.92
480588.40
5079415.24
517205.16
5046611.99
528207.10

Site Size (M²) M² Excavated
800
1.0
900
1.0
100
0.1
20000
102.0
1000
9.3
700
1.5
2400
13.5
900
1.0
8000
2.0
1400
9.9
1000
6.1
400
3.4
100
1.0
800
3.4
600
4.0
1400
1.0
1600
1.7
6690
116.0

217

Phase
Survey
Survey
Survey
Excavation
Excavation
Excavation
Excavation
Survey
Survey
Excavation
Excavation
Survey
Survey
Excavation
Excavation
Survey
Survey
Excavation

Component(s)
Single
Single
Single
Multiple
Multiple
Multiple
Multiple
Multiple
Multiple
Single
Single
Single
Single
Multiple
Multiple
Single
Multiple
Multiple

Table 41: (cont'd)
FS Number State Number

UTM Northing

UTM Easting Site Size (M²)

-

20MK169/457
20MK19

5080250.09
5079866.63

684990.60
677727.98

05-075

20MK22
20MK239
20MK24

5090459.89
5080867.08
5084786.32

601396.86 677988.97
667927.26

05-322
05-072
05-361
-

20MK261
20MK3/11
20MK334
20MK375

5099106.43
5077107.65
5097050.66
5101860.73

678506.73
687096.29 660679.55
682526.54

-

20MK51/82/99

5082298.16

676027.22 -

05-014
02-220/221

20MK53
20MK54
20MK58
20MK6/7
20MK61
20MK90
20ST1
20ST109/110

5081498.99
5080153.72
5102377.83
5081907.67
5080736.59
5088147.80
5086841.87
5105541.53

676954.41 678393.50
682069.88
670793.70 678093.36
663117.76
581924.02
540105.17

02-038
02-435
02-442
02-445

20ST14
20ST2
20ST227
20ST233
20ST262

5116496.45
5091422.10
5114163.12
5118031.92
5106413.48

544059.58
558893.78 531017.53
543956.56
540391.63

10000

0
4000
9900
200
1000

M² Excavated Phase
177.0
46.5

Excavation
Excavation

Multiple
Multiple

104.0
14.0

Excavation
Excavation

Multiple
Multiple

12.0
0.2
4.0

Excavation
Survey
Excavation

Multiple
Single
Multiple

-

-

51.0
73.0
14.0
113.0
2.3
10.0
4.8

Excavation
Excavation
Excavation
Excavation
Excavation
Excavation
Excavation

Multiple
Multiple
Single
Multiple
Multiple
Multiple
Single

4.0
7.8
1.4
11.5

Excavation
Excavation
Survey
Excavation

Single
Single
Single
Single

-

1415
200
1000
500
2787
650
500
300
1000
500

Component(s)

218

Reference
Andrews 2011; Martin and Perri 2011; Prahl
1986
Holman 1978; Martin 1985; Smith 1983
Hambacher personal communication; Martin
1982; McHale-Milner 1998
McHale-Milner 1998
Lynott 1974
Dunham at al. 1993; Rutter and Weir 1989
Drake, personal communication
Dunham and Branstner 1992
Mayry 1995
Branstner 1991; Fitting 1980; Fitting (ed.)
1976
Fitting and Lynott 1974; McHale-Milner 1998
Fitting and Clarke 1974
Fitting and Fisher 1975
Martin 1979
Fitting and Cushman 1974
Martin and Martin 1979
UMMA files; Martin 1985
Franzen 1983; Franzen 1987
Franzen 1979; 1998; Goltz 1992; Rutter and
Weir 1988
OSA files
Dunham et al. 1993; Goltz 1992
Franzen 1998; Goltz 1992
Dunham et al. 1993; Goltz 1992

State Number
20AR173/174
20AR310
20AR338
20AR348
20AR350
20AR358/386
20AR359
20AR398
20AR437
20AR495
20CH2
20DE167
20DE188
20DE296
20DE333
20DE4 (O/LW)
20DE4 (pHST)
20MK1 (Bois Blanc)
20MK1 (Juntunen)
20MK1 (Mackinac)
20MK169/457
20MK22
20MK24
20MK51/82/99
20MK53
20MK54
20MK6/7
20MK61
20MK90
20ST1
20ST109/110
20ST227

Count

Hearth

1
x
x
4
1
3
2
1
1
2
x
1
1
1
1
8
9
3
10
10
20?
x
1
?
1
15
?
10
2
?
1
1

1

1

Refuse
Dump

Refuse Pit

Storage
Pit

Basin

Roasting
Pit

1

Living
Floor

Small Pit

1

1

Net sinker
concentration

Midden

Clay
concentration

Dwelling

Animal
Burial

1
1

2

2
1
1
1

1

1
1
1
2
3
1
4
4
x

5
3

1
2

1
1
3
3
x

2

1
1

1
1

1

1

x = present
Table 42: Features

219

1

1
1
x

1
x

1

State No.
Site Name
20AR013
20AR310
20AR350
Popper
20AR358/386 Mather Lodge
20CH95
Bark Dock
20DE167
20DE4
Summer Island
20DE4
Summer Island
20DE4
Summer Island
20MK1
Juntunen
20MK1
Juntunen
20MK1
Juntunen
20MK1
Juntunen
20MK1
Juntunen
20MK1
Juntunen
20MK1
Juntunen
20MK1
Juntunen
20MK22
Scott Point
20MK22
Scott Point
20MK22
Scott Point
20MK24
Ferrier
20MK24
Tamlin
20MK53
20MK54
Beyer
20ST1
Ekdahl-Godreau
20ST1
Ekdahl-Godreau
Table 43: Chronometric Ages

Phase

Upper Mississippian
Upper Mississippian
Proto Historic
Bois Blanc
Bois Blanc
Juntunen
Juntunen
Mackinac
Mackinac
Mackinac
Mackinac
Late
Late
Mackinac

Upper
Lower

Type
14C
14C
14C
AMS
AMS
14C
14C
14C
14C
14C
14C
14C
14C
14C
14C
14C
14C
AMS
AMS
AMS
14C
14C
14C
14C
14C
14C

14C age BP Cal Range AD median prob AD
1130 ± 50
778-997
913
950±50
1013-1208
1096
1230±350
78-1409
789
790±40
1174-1281
1239
600±40
1294-1411
1349
1100±200
583-1280
925
660±100
1174-1442
1325
660±200
964-1666
1312
330±100
1410-1695
1575
890±75
1020-1265
1136
820±120
990-1328
1185
630±75
1263-1432
1345
620±75
1268-1433
1348
1050±75
808-1158
983
1225±75
662-904
801
870±120
945-1310
1130
890±120
932-1299
1146
860±40
1147-1261
1180
870±40
1118-1255
1170
1240±40
680-882
772
1020±90
855-1215
1017
900±85
994-1270
1128
310±85
1430-1691
1581
680±90
1179-1429
1312
870±120
945-1310
1146
1290±30
663-775
716

220

14C Early 14C Late # of ranges relative area
E
2
0.99
L
2
0.99
*
*
1
1.00
L
1
1.00
L
1
1.00
E
L
1
1.00
L
1
1.00
L
3
0.99
L
6
0.86
L
4
1.00
L
2
0.94
L
1
1.00
L
1
1.00
E
L
2
0.99
E
2
0.90
E
L
3
0.97
E
L
3
0.97
L
3
0.76
L
2
0.75
E
1
1.00
E
L
3
0.97
L
1
1.00
L
3
0.87
L
1
1.00
L
3
0.97
E
1
1.00

Sample Reference
Beta-46966
Beta-74502
WISC-2242
Beta-269591
Beta-214550
WISC-2244
M-2071
M-2072
M-2014
M-1140
M-1817
M-1188
M-1391
M-1141
M-1142
M-1815
M-1816
Beta-237014
Beta-237015
Beta-237016
N-1724
N-1725
N-1727
N-1726
M-2311
M-2312

Appendix C:

Additional Site Data

221

Additional Site Data
As noted in Chapter 3, the available reporting was reviewed for each of the LW sites in
the eastern UP. The archaeological site data were summarized in tabular form and is presented in
Appendices B through I. The analyses as presented by the original researchers were maintained
and no new analyses of the artifact assemblages were conducted. In some cases it was possible
to reorganize data presented in these reports to better address the needs of this study.
An attempt was made to present only the LW component of multicomponent sites, such
as at the Bark Dock site (20CH95), which includes Middle Woodland and LW components
(Dunham and Hambacher 2007) and Gete Odena site (20AR348) which includes Archaic,
Middle Woodland and Historic components in addition to a LW component (Dunham and
Branstner 1995; Robinson et al 1991; Skibo et al. 2004). The interpretations of the authors were
used for these sites.
Likewise, when multiple LW components were present on a site, an attempt was made to
differentiate them as separate components as well (e.g., the Juntunen and Summer Island sites
[Brose 1970; McPherron 1967]). This was not always possible and depended on the reporting of
the site.

As a result, a small number of sites (n=7) were characterized as LW locales based on

the recovery of LW artifacts, but where the LW component could not be differentiated from
other components on the site (20AR013, 20AR338, 20DE1, 20DE19, 20MK375,
20MK51/82/99, 20MK6/7).
Finally, an estimation of the LW component was derived from the original reporting for a
small number of multicomponent sites based on data in the original reports. For example, the
southern loci of site 20DE75 included LW artifacts and no diagnostic materials were noted in the
northern loci (Buckmaster 1983). Thus, data from the southern loci was used in this analysis.

222

Likewise, the Tamlin portion of 20MK24 was used in this study (Lynott 1974). LW features at
20CH2 were used to extrapolate diversity index values (Janzen 1968; Features 7-67, 31-67, 367). A similar process was used for 20CH6 using Features 18, 20, 21, 22, 23, 25, 26, 27, 30, 31,
37, 38, 39, and 40 (Branstner 1995).

223

Appendix D:

Chipped and Ground Stone Artifacts

224

State Number
Tool Count
20AR173/174
1
20AR245
10
20AR310
12
20AR330
1
20AR348
42
20AR353
6
20AR358/386
6
20AR359
14
20AR437
4
20AR495
7
20AR6
5
20CH171
5
20CH2
13
20CH32
2
20CH41
4
20CH6
26
20CH77
14
20CH86
1
20CH95
13
20DE108
4
20DE167
5
20DE188
16
20DE236
1
20DE296
15
20DE326
2
20DE333
1
20DE378
3
20DE4 (O/LW)
126
20DE4 (pHST)
178
20DE43
14
Table 44: Chipped Stone Tools

Formal Expedient Point
1
1
2
1
4
1
1
1
1
15
1
1
4
1
1
2
1
10
1
1
3
1
1
6
1
5
1
4
1
11
1
1
1
1
0
1
18
1
1
4
1
1
9
1
1
1
1
1
1
1
8
1
1
0
1
7
1
1
1
1
1
1
1
2
1
34
1
1
61
1
1
5
1

Scraper Biface Drill Knife Escraper Sscraper
1
1

1
1
1
1

1
1
1
1

1
1
1
1
1

1

1
1

1

1
1
1
1
1
1
1
1

1
1
1
1
1

1
1
1

1
1
1

1

1

1
1

1

1

1

1

1

1

1
1
1

1
1
1
1

1
1

1
1
1
1

225

Table 44: (cont'd)
State Number
Tool Count
20DE459
3
20DE50
5
20DE7
3
20DE75
9
20DE85
3
20DE93
4
20MK1
660
20MK1 (Bois Blanc)
59
20MK1 (Juntunen)
120
20MK1 (Mackinac)
93
20MK102
3
20MK159
3
20MK169/457
42
20MK19
28
20MK22
25
20MK24
2
20MK261
43
20MK54
77
20MK58
2
20MK61
39
20MK90
68
20ST1
40
20ST109/110
12
20ST227
2
20ST233
1
20ST262
1

Formal Expedient Point
1
1
1
2
1
1
3
1
1
1
3
1
4
1
443
1
1
28
1
1
70
1
1
63
1
1
3
1
1
1
7
1
1
7
1
1
10
1
1
1
31
1
1
24
1
1
2
1
10
1
1
15
1
1
21
1
1
4
1
1
1
1
1
1
1
1

Scraper Biface Drill Knife Escraper Sscraper
1
1

1
1

1
1
1
1
1
1

1
1

1

1
1

1
1
1
1
1
1
1
1
1
1
1

1
1
1
1
1

1
1

1

1

1

1

226

1

1
1
1
1
1
1
1
1
1
1
1

1
1
1
1

State Number Count
20AR173/174
0
20AR210
119
20AR245
382
20AR310
553
20AR348
851
20AR353
159
20AR358/386
1378
20AR359
761
20AR398
144
20AR400
56
20AR406
0
20AR435
0
20AR437
81
20AR495
632
20AR6
192
20CH171
44
20CH238
65
20CH27
24
20CH32
23
20CH41
53
20CH43
19
20CH433
1
20CH46
3
20CH492
0
20CH77
68
20CH86
8
20CH95
137
20DE108
77
20DE167
4281
Table 45: Debitage

Reduction Sequence (%)
First
24
25.3
33
20.5
55
25.9
4.3
15.3
10.4
-

Second Third
23.6
3
31
4.7
36.4
0.8
34.4
16.9
14
4
17.3
35.8
43.5
21.7
54.8
8.8
62.4
18.2
-

Raw Material (%)
Chert
99.2
61.8
86.6
28.6
27
2
13
57
16.1
91.4
29
78.6
98.5
95.8
100
62.5
93.4
97.4
96.8

227

Quartz Quartzite
0.8
0
0
36
4.7
8.7
51.1
20
54
19
93
4
24
63
16
27
12.5
67.9
6.2
1.2
41
30
15.6
5.2
0
1.5
0
4.2
0
0
37.5
0
5.1
1.5
2.6
0
0.8
2.4

Table 45: (cont'd)
State Number
20DE188
20DE236
20DE294
20DE296
20DE326
20DE378
20DE4 (O/LW)
20DE4 (PHST)
20DE43
20DE459
20DE50
20DE7
20DE75
20DE85
20DE93
20MK159
20MK169/457
20MK19
20MK24
20MK261
20MK334
20MK54
20MK58
20MK61
20MK90
20ST1
20ST109/110
20ST14
20ST227

Count
367
81
4
258
69
20
12150
14900
642
167
35
204
241
48
737
25
152
876
74
1982
8
371
5
338
888
453
195
35
98

Reduction Sequence (%)
First
19.7
17.6
5.9
100
28.2

Second Third
34.4
9.4
32.9
12.9
21.8
24.9
0
0
30.8
12.8

Raw Material (%)
Chert
95
63.6
100
92.5
94.1
95
95.9
22.7
88.6
97.9
93.8
100
99
97.6
100
98.4
99.7
86.7
74.3
19.3

228

Quartz Quartzite
0
5
36.4
0
0
0
3.8
1.3
2.9
2.9
0
5
0.6
3
7
69.8
5.7
5.7
0.4
0.8
0
6.3
0
0
2.1
2.1
0
0
0
1.3
12.8
0.5
14.3
11.4
15.3
61.2

Table 45: (cont'd)
State Number
20ST233
20ST262

Reduction Sequence (%)
Count
First
28
148
11.3

Raw Material (%)
Second Third
Chert
85.7
33.1
16.2
77.7

229

Quartz Quartzite
7.1
7.1
14.9
8.8

Ground Stone
Mano
Count
20AR310
2
20AR348
5
20AR358/386
2
20AR437
1
1
20AR495
14
20AR6
2
20CH95
5
20DE296
1
20DE333
1
20DE4 (O/LW)
6
2
20DE4 (pHST)
4
20DE43
2
20MK1
31
20MK1 (Bois Blanc)
1
20MK1 (Juntunen)
4
20MK1 (Mackinac)
2
20MK19
2
20MK22
25
6
20MK261
3
20MK90
1
20ST1
1
20ST109/110
2
20ST227
1
20ST262
1
1
Table 46: Ground Stone Tools
State Number

Hammer
Stone
2
2
1
2

Anvil Stone Net Sinker

Celts

Abrader Unclassified

Drilled Slate Adze Pestle

3
1
1

11
1
4

1

1

1
1
2
2
2
20
1
1
1
2
9
1

2
1

1

1

1

3
3

1
9

1
2

1
1
1
1

1

230

Appendix E:

Ceramics

231

State Number Sherds # MNV #
20AR210
1
1
20AR245
8
1
20AR310
4
1
20AR330
1
20AR348
671
4
20AR350
7
1
20AR353
2
2
20AR358/386
269
1
20AR359
163
3
20AR398
2
1
20AR400
126
2
20AR406
3
1
20AR435
91
1
20AR437
160
9
20AR495
6
1
20AR6
2
1
20CH171
2
1
20CH2
1089
58
20CH238
408
3
20CH27
1
20CH32
4
2
20CH41
6
1
20CH43
101
3
20CH433
1
1
20CH45
9
1
20CH46
353
1
20CH492
1
1
20CH6
136
20CH77
51
3
20CH86
1
1
20CH95
577
13
20DE106
2
20DE108
71
3
20DE17
1
20DE188
97
1
20DE236
11
3
20DE294
2
1
20DE296
466
17
20DE326
5
3
20DE333
p
1
20DE378
9
3
Table 47: Baseline Ceramic Data

Minimum Number of Vessels
No. Types
ELW
LLW
Untyped
1
1
1
1
1
1
1
1
4
1
2
1
1
1
2
1
1
2
1
2
2
1
1
1
1
2
1
1
1
1
4
0
9
1
1
1
1
1
1
1
4
51
7
3
3
1
1
2
1
1
1
1
1
2
1
2
1
1
1
1
1
1
1
1
8
48
88
64
3
3
4
2
1
1
2
1
3
2
1
1

13

6

1

2

1
1

2

232

17
1
1
1

1
2
1
4
2
3

Table 47: (cont'd)
State Number Sherds #
20DE4 (O/LW)
538
20DE4 (pHST)
66
20DE43
9
20DE50
6
20DE7
14
20DE75
162
20MK1
101447
20MK1 (Bois Blanc) 20MK1 (Juntunen) 20MK1 (Mackinac) 20MK102
20MK159
14
20MK169/457
224
20MK19
245
20MK22
20MK239
20MK24
117
20MK261
1773
20MK3/11
1
20MK334
100
20MK375
20MK53
41
20MK54
1557
20MK58
123
20MK61
327
20MK90
246
20ST1
609
20ST109/110
91
20ST14
88
20ST233
18
20ST262
1

MNV #
16
13
2
1
2
2
1656
142
269
438
12
2
27
3
195
3
19
8
1
1
1
5
19
1
10
2
8
2
3
1
1

Minimum Number of Vessels
No. Types
ELW
LLW
Untyped
4
16
5
13
1
2
2
1
1
2
1
1
2
1
1
11
*
*
463
*
*
*
4
8
1
3
1
2
13
9
16
2
3
x
8
6
1
3
3
2
12
2
8
1
1
1
1
2
6
1
4
1
2
2
2
1
1

3
1
2

1
6
1
4

4
8

7
1
2

1
1
1

5

1

233

State Number

Juntunen Bois Blanc Mackinac

20AR210
20AR245
20AR310
20AR330
20AR348
20AR350
20AR353
20AR358/386
20AR359
20AR398
20AR400
20AR406
20AR435
20AR437
1
20AR495
20AR6
20CH171
20CH2
2
4
51
20CH238
20CH27
20CH32
1
20CH41
20CH43
1
20CH433
20CH45
1
20CH46
1
20CH492
1
20CH6
3
17
20CH77
1
20CH86
20CH95
1
20DE106
20DE108
20DE17
1
20DE188
20DE236
20DE294
20DE296
0
0
0
20DE326
20DE333
20DE378
Table 48: Late Woodland Ceramic Types (MNV)

Sand
Point

Heins
Pt. Sauble Madison
Creek-like

Lakes
Phase

Miniature Pine River Traverse

Macomb- Spring
Algomalike
Creek-like
like

Waynelike

Blackduck

1

1

1

1

1
1

1

1

13
1
5
1

12

234

6

9

2

Table 48: (cont'd)
State Number

Juntunen Bois Blanc Mackinac

20DE4 (O/LW)
20DE4 (pHST)
20DE43
20DE50
20DE7
20DE75
20MK1
20MK1 (Bois Blanc)
20MK1 (Juntunen)
20MK1 (Mackinac)
20MK102
20MK159
20MK169/457
20MK19
20MK22
20MK239
20MK24
20MK261
20MK3/11
20MK334
20MK375
20MK53
20MK54
20MK58
20MK61
20MK90
20ST1
20ST109/110
20ST14
20ST233
20ST262

Sand
Point

Heins
Pt. Sauble Madison
Creek-like

Lakes
Phase

Miniature Pine River Traverse

Macomb- Spring
Algomalike
Creek-like
like

Waynelike

Blackduck

1
309

138

631

2
x
22

7
2
7
x
42

1
2
48
3

15

21

1

55

1
2

1
x

1

2

3
4

1
2
1
3

2

1

1

7
4
4

3

1

1
2

235

Oneota Related
Grand
State Number
Carcajou Koshkonong Pt. Detour
River
20AR210
20AR245
20AR310
20AR330
20AR348
20AR350
20AR353
20AR358/386
20AR359
20AR398
20AR400
20AR406
20AR435
6
1
20AR437
20AR495
20AR6
20CH171
20CH2
20CH238
20CH27
20CH32
20CH41
20CH43
20CH433
20CH45
20CH46
20CH492
20CH6
20CH77
20CH86
1
20CH95
20DE106
20DE108
20DE17
20DE188
20DE236
Table 49: Oneota-Related and Other Ceramic Types (MNV)

Lake
Winnebago

Delta

Bay de
Noc

Garden

Summer
Island

Untyped
Oneota

Other
Rameylike

Iroquoian

1

1
1

22
1

1

236

Table 49: (cont'd)
Oneota Related
Grand
State Number
Carcajou
River
20DE294
0
0
20DE296
20DE326
20DE333
20DE378
4
20DE4 (O/LW)
20DE4 (pHST)
20DE43
20DE50
20DE7
20DE75
20MK1
20MK1 (Bois Blanc)
20MK1 (Juntunen)
20MK1 (Mackinac)
20MK102
20MK159
1
1
20MK169/457
20MK19
20MK22
20MK239
20MK24
20MK261
20MK3/11
20MK334
20MK375
20MK53
20MK54
20MK58
20MK61
20MK90
20ST1
20ST109/110
20ST14
20ST233
20ST262

Koshkonong

Pt. Detour

Lake
Winnebago

Delta

Bay de
Noc

Garden

Summer
Island

Untyped
Oneota

Other
Rameylike

Iroquoian

1
1
1
3

3

6
1

4

3

3

2

1
42

2

1

5

67

1

2
1
7
1
2
1

237

15

Appendix F:

Other Artifacts

238

State Number
20AR400
20CH2
20CH6
20CH95
20DE188
20DE4 (O/LW)
20DE4 (pHST)
20MK1
20MK1 (Bois Blanc)
20MK1 (Juntunen)
20MK1 (Mackinac)
20MK169/457
20ST1

Ceramic
Pipe

Stone
Pipe

1
1
1
1

Bone
Artifacts

Bone
Awl

Bone
Point

1

Bone Bone Netting Bone
Bone
Harpoon
Needle
Chisel Needle

Bone Copper Copper Copper
Tubes Artifacts Awl
Bead

1

6
1

Copper Copper Copper
Copper
Copper Copper
Scrap
Pin
Ring Point/Cone Effigy
Knife

5
1

1

1
12
65
7
9
10
2
p

118
5
21
22

5
69
5
18
20

3
1

4
9

31

3
2

p

1

7

3
11
90
4
22
1
3

p = present

Table 50: Other Artifact Types

239

3
3
57

3
25

4

1
8

1

1

Appendix G:

DUI Rank

240

State Number
DUI
DUIrev DUI Rank
20AR173/174
Limited
1
1
20AR210
Limited
1
1
20AR245
Limited
9
1.2
20AR310
Limited
35
2.9
20AR330
Limited
4
4
20AR348
Extended
168
7
20AR350
Limited
1
1
20AR353
Intermediate
24
5.7
20AR358/386
Limited
25
2.2
20AR359
Limited
29
1.7
20AR398
Limited
1
1
20AR400
Limited
2
2
20AR406
Limited
1
1
20AR435
Limited
1
1
20AR437
Intermediate
52
8
20AR495
Intermediate
126
15.8
20AR6
Intermediate
40
20
20CH171
Intermediate
10
10
20CH2
Intermediate
207
1
20CH238
Limited
3
1.5
20CH27
Limited
1
1
20CH32
Limited
6
3.3
20CH41
Limited
5
5
20CH43
Limited
3
3
20CH433
Limited
1
1
20CH45
Limited
1
1
20CH46
Limited
1
1
20CH492
Limited
1
1
20CH6
Extended
770
7.5
20CH77
Limited
14
1
20CH86
Limited
4
2.7
20CH95
Extended
168
12.4
20DE106
Limited
2
2
20DE108
Limited
8
4
20DE167
Limited
2
1
20DE17
Limited
1
1
20DE188
Intermediate
36
5.9
20DE236
Limited
3
1
20DE294
Limited
1
1
20DE296
Intermediate
125
36.3
20DE326
Limited
8
2
20DE333
Intermediate
9
9
Table 51: Diversity Use Index (DUI) Rank

241

Table 51: (cont'd)
State Number
20DE378
20DE4 (O/LW)
20DE4 (pHST)
20DE43
20DE459
20DE50
20DE7
20DE75
20DE85
20DE93
20MK1 (Bois Blanc)
20MK1 (Juntunen)
20MK1 (Mackinac)
20MK102
20MK159
20MK169/457
20MK19
20MK22
20MK239
20MK24
20MK261
20MK3/11
20MK334
20MK53
20MK54
20MK58
20MK61
20MK90
20ST1
20ST109/110
20ST14
20ST227
20ST233
20ST262

DUI
10
448
702
36
1
9
15
6
6
12
680
1715
3024
30
6
170
84
1840
3
80
294
1
1
5
258
9
40
108
84
40
3
4
4
9

DUIrev
5.9
3.9
6
2.8
1
3.6
15
1
2.4
6
5.5
7.5
6.9
2.5
1.4
1
1.8
17.7
1
5.7
24.5
1
1
1
3.5
1
1
46.5
8.4
8.3
1
1
2.9
1

DUI Rank
Intermediate
Intermediate
Extended
Limited
Limited
Limited
Intermediate
Limited
Limited
Intermediate
Extended
Extended
Extended
Limited
Limited
Intermediate
Limited
Extended
Limited
Intermediate
Extended
Limited
Limited
Limited
Intermediate
Limited
Limited
Intermediate
Intermediate
Intermediate
Limited
Limited
Limited
Limited

242

Appendix H:

Fauna

243

11 sites 20AR359 20AR359 20DE75 20DE75 20MK90 20MK90 20CH95 20CH95
Taxon
MAMMALS
Beaver, Castor canadensis
Black bear, Ursus americanus
Canis sp., wolf/dog
Chipmunk (eastern), Tamias striatus
Porcupine, Erethizon dorsatum
Marten, Martes americana
Meadow jumping mouse, Zapus hudsonicus
Mink, Mustela vison
Moose, Alces alces
Muskrat, Ondatra zibethicus
River otter, Lutra canadensis
Short-tailed shrew, Blarina brevicauda
White-tailed deer, Odocoileus virginianus

BIRDS
Bald eagle, Haliateetus leucocephalalus
Common goldeneye (duck), Bucephala
clangula
Common loon, Gavia immer
Herring gull, Larus argentatus
Oldsquaw, Clangula hyemalis
Passenger Pigeon, Ectopistes migratorius
Red-tailed hawk, Buteo jamaicensis
Ring-necked duck, Aythya collaris
Eggshell fragments

NISP

NISP

MNI

x (7)
x (2)
x (1)
x (1)
x (2)
x (2)

NISP

MNI

14

1

NISP

MNI

20AR348 20AR348 20AR348 20AR348
20ST109/ 20ST109/
20MK261 2MK261 20DE188 20DE188 20MK24 20MK24
20DE296 20DE296 20AR437 20AR437
(1990)
(1990)
(1994)
(1994)
110
110
NISP
MNI
NISP
MNI
NISP
MNI
NISP
MNI
NISP
MNI
NISP
MNI
NISP
MNI
NISP
MNI
1
4

1
1

3
1

1
1

25

3

3

1

44

3

8
14

2
2

2
2
2

1
1
1

x (3)

3

1

x (1)
x (3)
x (3)
x (2)
x (1)

1
12
2

1
1
1

2

6

93

9

1

1
1

1

7
3

2

2
2

2

3

1

x (1)

1

1

x (1)

1

1

x (1)
x (1)
x (1)
x (1)
x (1)
x (1)
x (1)

4

1

20

-

6

1

1

1

41

8

2

1

18
2
3

1
2
3

2

2

2
20

1
3

112

7

5

3

1

1

1

x (4)

REPTILES
Painted turtle, Chrysemys picta
Snapping turtle, Chelydra serpentina

x (3)
x (1)

AMPHIBIANS
Frog, Rana sp.

x (1)

FISH
Bass (smallmouth), Micropterus dolomieui

MNI

1

1

2

1

1

1

Lake Sturgeon, Acipenser fulvescens
Lake trout, Salvelinus namaycush
Northern pike, Esox lucius
Sucker (longnose), Catostomus catastomus

x (8)
x (2)
x (4)

1

x (3)

1

Sucker (white), Catostomus commersoni
Sucker, Catostomidae
Sunfish, Centrarchidae
Walleye, Stizostedion vitrium
Whitefish (lake whitefish), Coregonus
clupeaformis
Whitefish fam., Coregoninae sp.
Yellow perch, Perca flavescens

x (1)
x (5)
x (1)
x (2)

4

1

2

2

2

1

1

1

1

-

26

3

1

5

1

1

1

3
p

1
-

243

1

1

1

2

2

-

1
1

1

x (1)

1

1

x (3)
x (1)

2

-

1

BIVALVES
Fatmucket, Lampsilis siliquoida
x (1)
cf. White heelsplitter, Lasmigona complanata
x (1)
Table 52: Fauna from sites within the HNF (Identified Species Only)

244

3

1

1

1

3

3
1

1

1
1

1

1

1

1

1

2

1

1

1

1

4
2

x (1)
1

13

2

1

1

x (1)
x (1)

1

1

x (2)

Burbot, Lota lota
catfish/bullhead, Ictalurus sp.
Freshwater drum, Aplodinotus grunniens

1

1
3

1

3

1
2

1

1

43

-

2

9 sites (16
components)
Taxon
MAMMALS
Beaver, Castor canadensis
Black bear, Ursus americanus
Canis sp., wolf/dog
Caribou, Rangifer species
Chipmunk (eastern), Tamias
striatus
Deermouse, Peromscus
maniculatus
Elk, Cervis canadensis
Fisher, Martes pennanti
Fox, Vulpes species
Marten, Martes americana
Moose, Alces alces
Muskrat, Ondatra zibethicus

20MK61

20MK457 (LW
Mixed)

20MK457 (LW
Mixed)

20MK457 (Bois
Blanc)

20MK457 (Bois
Blanc)

20MK457 (Early)

20MK457 (Early)

20CH6

20CH6

20ST1

20ST1

20DE4 (O/LW)

20MK1

20CH238

NISP

NISP

MNI

NISP

MNI

NISP

MNI

NISP

MNI

NISP

MNI

MNI

MNI

NISP

NISP

NISP

NISP

NISP

NISP

NISP

NISP

NISP

x (15)
x (6)
x (12)
x (7)

5
1
4

68
2
81
12

4
1
7
2

4

1

5

2

x

x
x

8

15

32

2
1

1

40
3

5
1
6
1

25

32
2

42
1
2

2

2
1

4
1
1
1

18

14
2

468
1
259
1

20
2

62
1
66
6

x (5)

4

1

1

2

3
3
18
1

2
1
2
1

2

1

2

2

2

1

3

1

Porcupine, Erethizon dorsatum

x (4)

5

2

1

Rabbit?, Lepus
Red squirrel, Tamiasciurus
hudsonicus
River otter, Lutra canadensis
Skunk, Mephitis mephitis

x (2)
x (1)
x (3)
x (1)

1

Snowshoe hare, Lepus americana

x (6)

1

BIRDS
American bittern, Botaurus
lentiginosus
Bald eagle, Haliateetus
leucocephalalus
Blue jay, Cyanocitta cristata
Canada goose, Branta canadensis

2

20MK54

20MK1 (Mackinac)20MK1 (Bois Blanc)20MK1 (Juntunen)

8
1

1

x
1
1
1
3

1

1

1
1
1

2
2

2

1

3

2
x

1
1

18

1
1

7

2

1
1

1

5
1

x

x

1

1

3

2

1

6

2

3
4

2

1

1

4

1
7

2

13

13

4

x (2)

7

x (1)

3

2

2

1

4

10
2

16
2
1

5

2

20

27

4

1

1

6

1

1

1

x (1)

Wild turkey, Meleagris gallopavo

2

1

1

2

x (3)

x (1)

3

1

x (6)

Saw whet owl, Aegolius acadicus

2

1

x (1)

x (3)

12

1

1
3

x (4)

Ruffed Grouse, Bonasa umbellus

3

1

x (1)

x (1)

3

1
1

6

x (1)

Herring gull, Larus argentatus

6

1

1

x (10)

x (2)
x (1)
x (8)
x (1)
x (1)

1

x

x (2)

Cuckoo, Coccyzus sp.
Common loon, Gavia immer
Crow, Corvis brachyrhynchos
Goshawk, Accipiter sp.

Passenger Pigeon, Ectopistes
migratorius
Raven, Coris corax
c.f. Red-shouldered hawk, Buteo
lineatus
Ringbilled gull, Larus
delawarensis

5

10

x (1)
x (1)
x (1)
x (3)
x (4)
x (9)
x (4)

Vole (red-backed), Myodes
species
White-tailed deer, Odocoileus
virginianus
Woodchuck, Marmota Mona

20DE4 (PHST) 20MK22 (Mackinac)
20MK22 (Bois Blanc)20MK22 (Juntunen)

1

1

1

1

1

1

1

1
5

5

REPTILES
Box turtle, Terrapene carolina

x (3)

Painted turtle, Chrysemys picta

x (5)

Snapping turtle, Chelydra
serpentina
Blandings turtle, Emydoidea
blandingii

1

1

x (5)

17

2

x (1)

AMPHIBIANS
Toad, Bufo sp.

x (1)

Catfish/bullhead, Ictalurus sp.
Channel catfish, Ictalurus
punctatus
Freshwater drum, Aplodinotus
grunniens

x (7)

469
2

1

2

4

1

23

33

x (11)
x (3)
x (7)

13

3

x?
2

1

14

4

1

1

1

7

2

9
1

1
3

2

1

1

3

28

5

1

1

Salmonidae
Northern pike, Esox lucius
Sucker (longnose), Catostomus
x (4)
catastomus
Table 54: Fauna from Sites outside the HNF (Identified Species Only)

10

3

2

1

6

2

4?

4

2

2

1

1

2

2

2

1

1
5

2

40

8

1

x (3)

x (12)

5

13

3

x?

x (1)

Lake Sturgeon, Acipenser
fulvescens

11

2

1

x (2)

x (2)

1

5

1

x (2)

x (3)

1

10

x (4)

Gar (longnose), Lepistosteus
osseus

Lake trout, Salvelinus namaycush

1

x (7)

Wood turtle, Glyptemys insculpta

FISH
Bass (smallmouth), Micropterus
dolomieui
Bass (largemouth), Microterus
salmoides
Bullhead (brown), Ameiurus
nebulosus
Burbot, Lota lota

2

1

6

28

x

5
9

3

1

1

x

29

39

9

3

5

17

87
2?

193
5?

65
4?

7

1

3

14

1

15

221

17

26

264

22

2

2

26

1

20

2
10

245

1

2

9

2

x

Table 54: (cont'd)
Taxon
Sucker (redhorse), Moxostoma
carinatum
Sucker (white), Catostomus
commersoni
Sucker, Catostomidae?
Walleye, Stizostedion vitrium
Whitefish (lake whitefish),
Coregonus clupeaformis

9 sites (16
components)

20MK61
NISP

20MK457 (LW
Mixed)
NISP

20MK457 (LW
Mixed)
MNI

20MK457 (Bois
Blanc)
NISP

20MK457 (Bois
Blanc)
MNI

20MK457 (Early)

20MK457 (Early)

20CH6

20CH6

20ST1

20ST1

20DE4 (O/LW)

NISP

MNI

NISP

MNI

NISP

MNI

MNI

20DE4 (PHST) 20MK22 (Mackinac)
20MK22 (Bois Blanc)20MK22 (Juntunen)
MNI

NISP

NISP

NISP

20MK54
NISP

x (3)
x (5)
x (4)
x (12)

1

4

1

1

1

x (9)

7

2

1

1

2

2

Whitefish fam., Coregoninae sp.

x (6)

215

11

27

2

79

3

Yellow perch, Perca flavescens

x (3)

36
26

3
3

23

3

3

1

4

x

3

20

2

5
1

1
1

4

2

8

BIVALVES
7

Fatmucket, Lampsilis siliquoida

x (1)

Fluted-shell, Lasmigona costata

x (2)

1

1

Plain Pocketbook, Lampsilis
cardium

x (2)

2

2

Pink heelsplitter, Potamilus alatus

x (2)

4

2

Actinonaias carinata
Spike, Elliptio dilatus
Amblema costata

x (1)
x (3)
x (1)

10

5

3
2
17

246

20MK1 (Mackinac)20MK1 (Bois Blanc)20MK1 (Juntunen)

20MK1

20CH238

NISP

NISP

NISP

NISP

NISP

6

1

4

11

6

10

2

24

5

19

16
48

1

125

31

30

186

10

4

14

State Number

Setting ELW LLW

20AR348
C
20AR359
C
20AR437
I
20CH238
I
20CH6
C
20CH95
C
20DE188
I
20DE296
C
20DE4 (O/LW)
C
20DE4 (pHST)
C
20DE75
I
20MK1 (Bois Blanc)
C
20MK1 (Juntunen)
C
20MK1 (Mackinac)
C
20MK169/457 (Bois Blanc)
C
20MK169/457 (Early)
C
20MK169/457 (Mixed LW)
C
20MK22 (Bois Blanc)
C
20MK22 (Juntunen)
C
20MK22 (Mackinac)
C
20MK24
C
20MK261
C
20MK54
C
20MK61
C
20MK90
C
20ST1
C
20ST109/110
I
Table 54: Faunal Diversity Index

E
E
E
E
E
E
E
E
E
E
E
E
E
-

L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L

Diversity
Mammals
6
0
4
1
6
1
1
11
3
5
0
8
8
11
6
5
12
8
9
5
0
3
3
4
1
3
3

Diversity
Birds
3
0
0
0
1
0
0
3
0
0
0
6
8
7
0
0
3
4
2
5
1
1
1
0
0
0
0

247

Diversity
Reptiles
0
0
0
0
1
0
0
1
0
0
0
2
1
1
1
1
3
3
3
3
0
0
0
0
0
2
2

Diversity
Fish
4
1
2
0
9
1
3
9
3
2
1
13
7
15
2
4
7
9
6
7
0
6
3
5
1
3
3

Total
Faunal
13
1
6
1
17
2
3
24
6
7
1
29
24
34
9
10
25
24
20
20
1
10
7
9
2
8
8

Diversity (total fauna
x variety)
39
2
12
1
68
4
6
96
12
14
1
116
96
136
27
30
100
96
80
80
1
30
21
18
4
24
24

Appendix I:

Flora

248

12 Sites (13
components)
Fruits
Crategus sp. (Hawthorn)

1

Arctostaphylus uva-ursi (bearberry)

1

Frageria/Potentilla sp.

(Strawberry)/(Cinquefoil)
Prunus sp.(Cherry)

20MK1 20CH6 20DE4 (O/LW)

20DE4 (Phst) 20DE75 20DE296 20DE326 20AR437

20AR348 20AR358 20AR359 20MK24 20ST109/110

2
x

1

15

5
1

x

Prunus virginiana (Choke Cherry)
Prunus pennsylvanica (pin cherry)

2

x

Prunus americana (Wild Plum)
Prunus nigra (canada plum)
Rubus sp. (Raspberry)
Rhus sp. (Sumac)
Sambucus sp. (Elderberry)

1
1
2
1
2

Sorbus americana (mountain ash)

1

Vaccinium sp. (blueberry)
Vitis sp. (Grape)

1
2

x
2

x

4

x

3 (27)
3

x
x
x
x

1

6

2
1 (4)

28

3 (2)
x

1

Other Seeds

cf. Asteraceae (Aster Family)

1

cf. Cornus sp. (Dogwood)
Dentaria laciniata (pepper root)
cf Fabicae (pea/legume family

1
1
1
3

Galium sp. (Bedstraw)

Scirpus spp. (bullrush)
cf. Viola sp (Violet)
Nutshell
Corylus sp. (Hazelnut)
Fagus grandifolia (beechnut)
Juglandaceae (Walnut Family)
Quercus sp. (Acorn)

1
1
x

1
x

6
2
2

17

1
2

5
1
1
6

2

x
x
x

5

7

8

2

1
115

2

2

2

6

31

1

Nutmeats
Quercus sp. (Acorn)

2

Domesticates
Zea mays Kernel Count
Cucurbita pepo

3
1

x

7

15

28

Table 55: Botanical Remains (PEB Identified/Carbonized Only)

249

12 Sites (13
components)

Table 55: (cont'd))

20MK1 20CH6 20DE4 (O/LW)

Cultigens & Grains
Chenopodium sp. (Chenopod)
Polygonum spp. (Knotweed)
Zizania aquatica (Wild Rice)

2
1
1

x

Other Plant Remains
cf., Aquatic Tuber

2

x

Betula papyrifera (paper birch) bark

1

x

Diversity Seed/Nut Remains

17

20DE4 (Phst) 20DE75 20DE296 20DE326 20AR437

20AR348 20AR358 20AR359 20MK24 20ST109/110

3
2
1

x

16

1

3

2

x = present

250

3

1

4

2

2

1

1

5

Acer spp. (maple)
Acer saccarum (sugar maple)
Betula spp. (birch)
Carya spp. (hickory)
Conifer

No. of Sites 20MK54 20MK61 20DE296 20DE326 20AR437
x
4
13
1
5
2
3

6
4

x
x

2
12

Conifer Type A (cf., Tsuga/ hemlock)

2

14

Conifer Type B (cf., Larix/ tamarack)

1

Conifer Type C (cf., Abies/fir)
Fraxinus spp. (ash)
Fraxinus americanus (white ash)
Pinus spp. (pine)
Pinus resinosa (red pine)
Pinus strobus (white pine)
Quercus spp. (oak)

1
3
2
2
1
1
1

2
3
31

Quercus bicolar (swamp white oak)

1

Ulmus spp. (elm)

1

20AR348

1

20AR359 20MK24
6
x
x
2
x
28

1
x

x

4

1

1
13
1
x?

1

Table 56: Wood Charcoal Remains (PEB Identified/Carbonized Only)

251

Appendix J:

HNF Site Data

252

State No.
Setting
20AR013
Coastal
20AR173/174
Interior
20AR245
Interior
20AR310
Interior
20AR338
Coastal
20AR348
Coastal
20AR350
Coastal
20AR353
Coastal
20AR358/386
Coastal
20AR359
Coastal
20AR398
Coastal
20AR400
Coastal
20AR406
Coastal
20AR435
Coastal
20AR437
Interior
20AR495
Coastal
20AR6
Coastal
20DE106
Coastal
20DE108
Coastal
20DE167
Interior
20DE188
Interior
20DE236
Coastal
20DE294
Interior
20DE296
Coastal
20DE326
Interior
20DE378
Interior
20DE43
Interior
20DE459
Interior
20DE50
Interior
20DE75
Interior
20DE85
Coastal
20DE93
Interior
20ST109/110
Interior
20ST14
Interior
20ST227
Interior
20ST233
Interior
20ST262
Interior
20CH171/172
Interior
20CH2
Coastal
20CH32
Coastal
20CH433
Coastal
20CH492
Coastal
Table 57: HNF Site Data

HNF Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
West Unit
East Unit
East Unit
East Unit
East Unit
East Unit

Ownership
PIRO
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
Private
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
HNF
Private
HNF
HNF
HNF
HNF

253

Survey
NPS CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
UMMA CR Survey
Informant, confirmed
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
MDOT CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
SHPO Site File
UMMA CR Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey

Table 57: (cont'd)
State No.
Setting
20CH86
Coastal
20CH95
Coastal
20MK159
Coastal
20MK24
Coastal
20MK261
Coastal
20MK3/11
Coastal
20MK334
Interior
20MK375
Coastal
20MK58
Coastal
20MK6/7
Coastal
20MK90
Coastal

HNF Unit
East Unit
East Unit
East Unit
East Unit
East Unit
East Unit
East Unit
East Unit
East Unit
East Unit
East Unit

Ownership
HNF
HNF
HNF
HNF
HNF
HNF
HNF
Private
Private
Private
HNF

254

Survey
HNF CR Survey
HNF CR Survey
HNF CR Survey
SOM CR Survey
HNF CR Survey
MDOT CR Survey
HNF CR Survey
Informant, confirmed
Informant, confirmed
MDOT CR Survey
MDOT CR Survey

Appendix K:

Archaeological Surveys on the HNF

255

Archaeological Survey on the Hiawatha National Forest
Archaeological surveys and excavations have been carried out in the eastern UP for well
over 60 years, although the first concerted efforts were carried out by the University of Michigan
in the 1960s. These projects were primarily geared towards Great Lakes coastal areas and
provide the basic framework for our understanding of Woodland sequences in the region. These
projects included excavations at such sites as Summer Island (Brose 1970), Juntunen
(McPherron 1967), and Naomikong Point (Janzen 1968).
Direct archaeological information for LW sites in the interior of the eastern UP came
about largely as a result of federal guidelines, primarily the National Historic Preservation Act,
coming into effect in the late 1960s. This led to the US Forest Service establishing Heritage
Programs to manage National Forest lands. Michigan State University developed sensitivity
models and implemented surveys for the HNF in the late 1970s: Martin’s (1977) resource
management study; and Lovis’ (1979) field test of that study.
Martin’s (1977) study stratified a research universe according to natural environmental
variables. These were then tested by Lovis’ (1979) field survey to determine which variables
correlated with prehistoric site loci. The results suggested that specific surface geological
contexts, soil associations, and vegetation types are associated with prehistoric site locations, but
only when they are spatially associated with major water features (see also Buckmaster 1979;
Franzen 1983; Lovis 1976).
Lovis’ (1979) test was a systematic survey of a randomly generated set of 186 quartersection (160-acre) survey parcels (approx. 29,760 acres) of HNF. This entailed the
implementation of survey transects at a maximum interval of 25 yards (ca. 22.5 m) across each
of the subject quarter-sections. This unbiased, systematic survey identified 32 archaeological

256

sites, all of which were historic.
A second systematic survey was conducted by Soil Systems, Inc. (Dorwin et al. 1980).
This survey employed 60 foot (18.3 m) interval linear shovel test transects paralleling the long
axis of each survey area (the survey areas were not randomly generated, rather representing
locations of planned Forest Service activity). The 113 survey areas, spread throughout the HNF,
included an aggregate of about 51,000 ac, of which 14,300 ac were directly surveyed. The
survey recorded 34 archaeological sites, 6 of these were prehistoric. Each of the prehistoric sites
was found in close proximity to water. The two systematic surveys directly examined about
44,060 ac, which represents an approximately 5 percent sample of the HNF.
The first quantified study of prehistoric site location on the HNF was carried out by
Franzen (1983) who found that 52 of the 53 verified prehistoric sites on the West Unit of the
HNF were located within 100 m of water and the remaining site within 150 m of water. On the
East Unit, ten of the twenty verified prehistoric sites were situated within 200 m of a Great Lake
shoreline and nine of the remaining sites were located on Holocene shoreline features between
90 m and 600 m from an extant Great Lake shoreline. The last East Unit site was located within
20 m of a major river as well as on a mid Holocene beach terrace within 300 m of Lake
Michigan. Because the data included both random and non-random surveys (the latter including
the early, shoreline-focused work of the University of Michigan and Michigan State University),
there is a risk that much of the apparent high degree of correlation of prehistoric sites and water
is the result of a circular argument.
Despite the relationship between water and prehistoric archaeological sites identified by
the early archaeological studies, it was recognized that ethnographic and geomorphological
evidence suggested the presence of archaeological site types not directly associated with extant

257

sources of water (Franzen 1986; Martin 1977). The most sensitive locations potentially distant
from extant water were determined to include post-Pleistocene (Holocene) shoreline features,
bedrock outcrops where raw material for stone tool making might be found, and wetland edges.
The results of the two systematic surveys of the HNF and the other early studies, as well
as the results of similar systematic surveys conducted on the Nicolet National Forest in
Wisconsin and the Huron-Manistee National Forests in northern lower Michigan (Lovis et al.
1978; Ryden et al. 1983), led to the development of the archaeological survey methodologies
used for locating prehistoric Native American sites by the HNF.
The methodology for archaeological survey used by the HNF since 1984 has relied on a
stepwise, cultural landscape/landform-based survey and testing strategy that uses a combination
of surface reconnaissance and shovel testing techniques (Anderton et al. 1991; Dunham and
Branstner 1998; Dunham et al. 2010; Franzen 1986; Rutter and Weir 1985). Generalized areas
of increased Native American archaeological site sensitivity are minimally defined as habitable,
level, and well-drained surfaces lying within 300 m of riparian features and wetland edges, as
well as identifiable post-Pleistocene terraces, beaches, and strand lines.
Between 1984 and 1990 a stratified survey methodology was used in areas defined as
high sensitivity for prehistoric archaeological sites. In areas within 100 m of a major riparian
feature or on a post-Pleistocene terrace, shovel testing was conducted on a 15 m interval grid. A
30 m interval grid was used in areas 100 m to 300 m of such features. Similarly, 15 m and 30 m
transect zones were employed along wetland edges and bedrock outcrop areas. Shovel testing
was not carried out if the area within the shovel test zone was exposed bedrock, thin soil horizon,
wetland, or steeply sloped. Exposed unimproved road and trail surfaces were also routinely
examined for prehistoric artifacts regardless of their relationship to a water source.

258

Prehistoric sites encountered by these surveys were typically within 100 m of a major
riparian resource or associated with a post-Pleistocene beach ridge. A random examination of
the results of the 1987 and 1990 surveys place 29 of the 43 prehistoric sites recorded (67 percent)
within 100 m of an extant major riparian resource (Anderton et al. 1991; Rutter and Weir 1988).
Two additional sites are within 150 m of an extant source of water. Twelve sites (27.9 percent)
were recorded as 200 m or greater from an extant body of water with two of these sites each
located 1600 m from a water source. Eleven of the sites over 200 m from water, including the
two that were 1600 m from water, are situated on mid-Holocene shoreline features. The one that
is not on a Holocene lakeshore is 200 m from a major river, but is also situated along a known,
historic Native American trail (Anderton et al. 1991:203).
Beginning in the 1990s the methodology was modified and the emphasis was placed on
shovel testing within 150 m of riparian features and Holocene beaches. This step-wise approach
also placed more latitude in observing field conditions, allowing field identification of potential
sensitivity areas inside and outside of the 150 m zone as well as in relation to bedrock outcrops,
wetland edges, and a variety of Holocene shoreline features (Dunham and Branstner 1998;
Dunham et al. 2010).
Using these methods, five hundred and seventy eight archaeological sites with prehistoric
components have been found on the HNF as of 2009. Four hundred and seventy five of these
have been identified on the West Unit and the remaining 103 on the East Unit. Forty eight of
these include LW components with 36 in the West Unit and 12 in the East Unit. The HNF
surveys have identified prehistoric sites in both coastal and interior contexts which provide a
better balance of general site contexts than the coastal oriented University of Michigan surveys.
An inherent problem with stratified surveys, like those carried out on the HNF, is that

259

they don’t apply the same level of survey coverage uniformly across the landscape. This raises
the question as to whether archaeological sites were found only in places that were intensively
surveyed, and if sites were not found in areas that were not surveyed because they weren’t
surveyed. A case can be made for survey bias on the HNF, with an emphasis on intensive survey
coverage (shovel testing) focused along major riparian resources. This approach may have led to
a false correlation between water sources and the location of prehistoric archaeological sites.
Although this is a valid concern, recall that the two systematic surveys, a five percent sample of
HNF lands, only identified prehistoric sites in close proximity to water (Dorwin et al. 1980;
Lovis 1979; see also Lovis et al. 1978; Ryden et al. 1983). Further, later stratified survey
samples identified sites farther than 150 m from extant sources of water. The negative data
resulting from these surveys, although not conclusive, along with data from other survey efforts
in the northern Michigan and northern Wisconsin make a strong case for the relationship
between prehistoric archaeological site locations and proximity to sources of water.
More recently, a systematic archaeological survey was conducted immediately outside
the HNF study area in Menominee County which is also in the Upper Peninsula (Dunham et al.
2011). This survey directly investigated 693 acres of land through walkover reconnaissance and
shovel testing. This survey was stratified, with areas within 500 m of the Menominee River
shovel tested at 15 m interval (238 acres) and those areas greater than 500 m from the river
shovel tested at 20 m intervals (455 acres). Fourteen prehistoric sites were identified as a result
of this survey and five of these can be directly ascribed to the LW period. Thirteen of the
prehistoric sites, including all the LW sites, were found within 150 m of a river or stream. One
site was found about 2000 m from the Menominee River, but this site was situated on a welldefined terrace overlooking a wetland that appears to have once been a small lake or pond.

260

Although this survey was somewhat limited in the size of the survey area and used a stratified
methodology, it supports the premise that prehistoric sites in the Upper Peninsula are typically
found within 150 m of a riparian feature. Unfortunately, the results of this survey were
undoubtedly also biased by the premise (reflected in the survey zones) that most sites would be
near water. Furthermore, if there is a difference in site types and size for the near-water versus
interior sites, many of the small, short-term sites of the interior may have been missed using a
20-m interval (i.e., many of the interior sites may be significantly smaller than 20 m in diameter).
In a related issue, determining where more intense archaeological survey has been carried
out on the HNF may be a challenge. The HNF documents areas in their GIS files as either
surveyed or not surveyed with no indication of the intensity of the survey coverage (this is also
true in the HNF Survey Atlas). Some of the data concerning survey intensity is available in the
appendices of archaeological surveys completed by contractors, but not all previously completed
surveys are equally documented. Thus, it is not currently possible to definitively quantify which
areas of the forest have been intensively surveyed (shovel tested), moderately surveyed
(walkover reconnaissance), or not directly surveyed. The formal compilation of this data will be
required for future explorations of site distribution on the forest.

261

Appendix L:

Variables and Summary Statistics

262

Summary Statistics

LW Sites

Random Points

32

43

Mean

161.5

186.1

Median

172.7

180.0

St. Dev.

78.6

106.1

Minimum

0.1

0.0

Maximum

313.3

353.0

36

50

657.5

747.7

Median

619

751.5

St. Dev.

82.2

82.8

Minimum

582

594

Maximum

856

921

Slope (n=)

36

50

14.0

5.6

Median

6

3

St. Dev.

20.9

8.1

Minimum

0

0

Maximum

82

84

D. to Water (n=)

36

50

66.3

459.4

Median

60

355

St. Dev.

66.4

360.8

Minimum

0

0

Maximum

335

1591

Grow Days (n=)

36

50

113.8

112.6

Median

110

110

St. Dev.

13.8

11.1

Minimum

99

99

Maximum

140

140

Aspect (n=)

Elevation (n=)
Mean

Mean

Mean

Mean

Table 58: Variables and Summary Statistics

263

D. to
Elevation Habitat Pre 1800
Water
20AR173/174
12
256
856
3
4
20AR245
23
30
805
2
4
20AR310
15
0
827
2
3
20AR338
6
30
608
2
4
20AR348
4
42
611
2
4
20AR350
18
67
622
2
4
20AR353
75
30
611
2
4
20AR358/386
18
67
625
5
4
20AR359
5
67
620
2
4
20AR398
8
30
607
2
4
20AR400
8
60
609
2
4
20AR406
12
335
621
1
4
20AR435
2
67
607
2
2
20AR437
23
42
790
2
3
20AR495
15
84
618
5
5
20AR6
2
0
602
5
7
20DE106
0
30
583
2
5
20DE108
82
67
582
7
4
20DE167
1
30
591
5
1
20DE188
5
30
621
5
3
20DE236
0
189
585
2
3
20DE294
0
0
585
4
1
20DE296
0
67
584
7
5
20DE326
20
90
738
2
2
20DE378
2
90
583
5
3
20DE43
6
67
764
2
3
20DE459
26
42
714
2
3
20DE50
24
67
726
2
2
20DE75
0
30
600
2
2
20DE85
74
67
585
2
1
20DE93
0
30
586
4
1
20ST109/110
7
42
674
2
5
20ST14
2
90
738
2
3
20ST227
1
30
770
2
2
20ST233
6
60
743
2
3
20ST262
1
60
678
2
5
Table 59: Variables and Summary Statistics LW Sites
Site No.

Slope

264

Aspect
66.3
219.4
206.9
165.5
130.9
131.8
133.2
233.3
154.7
184.6
183.4
95.2
206.6
214.9
216.3
0.1
-1.0
199.5
116.6
280.7
-1.0
-1.0
90.0
259.2
180.0
225.0
116.3
60.9
-1.0
313.3
135.0
67.6
225.0
63.4
15.6
278.1

Growing
Wild Rice
Days
99
0
100
1
99
1
100
0
100
0
100
0
100
0
110
0
110
0
100
0
100
0
100
0
100
0
100
1
100
0
100
0
130
0
130
0
130
0
130
0
140
0
130
0
130
0
110
1
130
0
120
0
120
0
120
0
130
0
140
0
130
0
120
1
110
0
110
0
110
0
110
1

WU
D. to
Growing
Slope
Elevation Habitat Pre 1800 Aspect
Random
Water
Days
WU1
1
726
837
5
4
116
99
WU2
0
182
758
4
3
-1
100
WU3
34
150
747
5
4
207
110
WU4
0
926
707
5
7
180
110
WU5
0
1376
673
5
7
71
120
WU6
25
308
785
2
2
235
110
WU7
0
108
740
2
3
341
110
WU8
0
270
682
2
2
0
110
WU9
35
94
688
6
2
230
110
WU10
2
210
647
5
5
144
120
WU11
13
785
838
4
4
315
120
WU12
12
891
832
4
4
76
120
WU13
4
1135
838
3
4
27
110
WU14
8
540
781
4
5
250
110
WU15
7
361
785
5
2
297
110
WU16
17
254
786
2
2
75
110
WU17
0
30
797
5
5
-1
110
WU18
0
666
780
2
2
-1
120
WU19
3
90
647
4
4
161
120
WU20
1
534
773
4
4
281
120
WU21
6
660
697
0
5
272
120
WU22
0
349
767
4
4
135
120
WU23
1
342
751
5
7
168
120
WU24
3
517
744
5
7
101
120
WU25
0
305
747
5
5
-1
120
WU26
3
210
753
5
7
145
120
WU27
1
646
735
5
7
251
130
WU28
1
894
738
3
2
270
130
WU29
0
30
620
7
5
-1
120
WU30
0
469
758
4
3
-1
130
WU31
0
436
629
5
5
-1
130
WU32
0
807
594
5
5
161
130
WU33
1
152
613
5
5
153
130
WU34
2
60
594
5
3
270
140
WU35
4
192
623
4
4
14
99
WU36
5
577
921
4
4
248
99
WU37
4
150
866
4
4
273
99
WU38
14
465
887
4
4
198
99
WU39
11
120
731
4
4
349
99
WU40
3
94
876
4
4
341
99
WU41
2
941
908
4
4
0
100
Table 60: Variables and Summary Statistics Random Points

265

Wild
Rice
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0

Table 60: (cont'd)
WU
Slope
Random
WU42
10
WU43
3
WU44
12
WU45
4
WU46
3
WU47
1
WU48
10
WU49
13
WU50
0

D. to
Growing
Elevation Habitat Pre 1800 Aspect
Water
Days
270
802
4
4
17
110
67
726
4
4
51
110
502
719
4
4
353
99
379
664
4
3
220
100
1023
624
5
5
321
100
1591
829
5
7
90
100
342
820
4
4
325
110
152
752
4
4
179
100
590
778
4
4
90
100

266

Wild
Rice
0
0
0
0
0
0
0
0
0

Appendix M:

Pre-1800 Vegetation Coding

267

VEGCODE
COVERTYPE
31
GRASSLAND
41
MIXED HARDWOOD SWAMP
42
MIXED CONIFER SWAMP
51
LAKE/RIVER
51
LAKE/RIVER
51
LAKE/RIVER
52
LAKE/RIVER
72
SAND DUNE
333
PINE BARRENS
334
OAK/PINE BARRENS
413
ASPEN-BIRCH FOREST
414
MIXED HARDWOOD SWAMP
423
MIXED CONIFER SWAMP
744
EXPOSED BEDROCK
744
SAND DUNE
4111
BEECH-SUGAR MAPLE-HEMLOCK FOREST
4141
BLACK ASH SWAMP
4143
MIXED HARDWOOD SWAMP
4146
MIXED HARDWOOD SWAMP
4147
MIXED HARDWOOD SWAMP
4148
MIXED HARDWOOD SWAMP
4211
WHITE PINE-RED PINE FOREST
4212
WHITE PINE-RED PINE FOREST
4213
JACK PINE-RED PINE FOREST
4215
JACK PINE-RED PINE FOREST
4216
WHITE PINE-RED PINE FOREST
4219
WHITE PINE-MIXED HARDWOOD FOREST
4221
SPRUCE-FIR-CEDAR FOREST
4223
SPRUCE-FIR-CEDAR FOREST
4226
HEMLOCK-WHITE PINE FOREST
4227
HEMLOCK-WHITE PINE FOREST
4228
SUGAR MAPLE-HEMLOCK FOREST
4229
HEMLOCK-YELLOW BIRCH FOREST
4231
CEDAR SWAMP
4232
MIXED CONIFER SWAMP
4233
MIXED CONIFER SWAMP
4235
MIXED CONIFER SWAMP
4236
MIXED CONIFER SWAMP
4237
MIXED CONIFER SWAMP
4238
MIXED CONIFER SWAMP
6121
MUSKEG/BOG
6122
SHRUB SWAMP/EMERGENT MARSH
Table 61: Pre-1800 Vegetation Coding

268

Pre-1800
0
6
5
7
7
7
7
0
1
1
4
6
5
0
0
4
6
6
6
6
6
2
2
1
2
2
2
3
3
3
3
4
3
5
5
5
5
5
5
5
7
7

Table 61: (cont'd)
VEGCODE
COVERTYPE
6124
MUSKEG/BOG
6125
MUSKEG/BOG
6221
SHRUB SWAMP/EMERGENT MARSH
6222
SHRUB SWAMP/EMERGENT MARSH
6223
SHRUB SWAMP/EMERGENT MARSH
6224
SHRUB SWAMP/EMERGENT MARSH
6227
WET PRAIRIE
6228
SHRUB SWAMP/EMERGENT MARSH
6231
CEDAR SWAMP

269

Pre-1800
7
7
7
7
7
7
7
7
5

Appendix N:

Habitat Classification Coding

270

Numeric
10
61
62F
68
11C
11E
11F
242B
242D
242F
297B
297D
305B
12B
12D
12E
15A
109D
109F
176B
307B
307D
308B
308D
309B
309D
255D

Name
Beaches
Pits, sand and gravel
Udipsamments and Udorthents, nearly level to very steep
Pits, quarry
Deer Park sand, 0 to 10 percent slopes
Deer Park sand, 10 to 25 percent slopes
Deer Park sand, 25 to 60 percent slopes
Kalkaska sand, 0 to 6 percent slopes, severely burned
Kalkaska sand, 6 to 15 percent slopes, severely burned
Kalkaska sand, 35 to 70 percent slopes, severely burned
Rubicon sand, 0 to 6 percent slopes, severely burned
Rubicon sand, 6 to 15 percent slopes, severely burned
Wurtsmith-Meehan sands, 0 to 8 percent
Rubicon sand, 0 to 6 percent slopes
Rubicon sand, 6 to 15 percent slopes
Rubicon sand, 15 to 35 percent slopes
Croswell sand, 0 to 3 percent slopes
Rousseau-Dawson complex, 0 to 15 percent slopes
Rousseau-Dawson complex, 0 to 60 percent slopes
Croswell-Kinross complex, 0 to 6 percent slopes
Rubicon sand, 0 to 6 percent slopes, very deep water table
Rubicon sand, 6 to 15 percent slopes, very deep water table
Rubicon-Sultz complex, 0 to 6 percent slopes
Rubicon-Sultz complex, 6 to 15 percent slopes
Rubicon sand, 0 to 6 percent slopes, deep water table
Rubicon sand, 6 to 15 percent slopes, deep water table
Wallace sand, 1 to 15 percent slopes

290A

Namur-Ruse complex, 0 to 2 percent slopes, very rocky, very stony

298B
300F
310B
310D
310E
13B
13D
13E
16A
24B
25B
25D
31D
Table 62:

Wurtsmith-Deford complex, 0 to 6 percent slopes
Shelldrake-Duneland complex, 2 to 75 percent slopes
Kalkaska sand, 0 to 6 percent slopes, burned
Kalkaska sand, 6 to 15 percent slopes, burned
Kalkaska sand, 15 to 50 percent slopes, burned
Kalkaska sand, 0 to 6 percent slopes
Kalkaska sand, 6 to 15 percent slopes
Kalkaska sand, 15 to 35 percent slopes
Paquin sand, 0 to 3 percent slopes
Munising fine sandy loam, 1 to 6 percent slopes
Munising-Yalmer complex, 1 to 6 percent slopes
Munising-Yalmer complex, 6 to 18 percent slopes
Trenary silt loam, 6 to 15 percent slopes
Alger County

271

Habitat
0
0
0
0
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4

Table 62: (cont'd)
Numeric Name
Munising-Yalmer-Frohling complex, calcareous substratum, 1 to 6
35B
percent slopes
37B
Grand Sable loamy fine sand, 1 to 6 percent slopes
37E
Grand Sable loamy fine sand, 15 to 35 percent slopes
40B
Waiska cobbly loamy sand, 0 to 6 percent slopes, very stony
47C
Deerton-Au Train complex, 1 to 15 percent slopes
47E
Deerton-Au Train complex, 6 to 35 percent slopes
49B
Cookson fine sandy loam, 1 to 6 percent slopes
52B
Summerville fine sandy loam, 1 to 6 percent slopes
64B
Kiva fine sandy loam, 1 to 6 percent slopes
64D
Kiva fine sandy loam, 6 to 15 percent slopes
Ruse-Ensign-Nykanen complex, bedrock terrace, 1 to 20 percent
66D
slopes
Ruse-Ensign-Nykanen complex, bedrock terrace, 1 to 45 percent
66F
slopes
69B
Escanaba sand, 1 to 6 percent slopes
Deerton-Tokiahok-Trout Bay complex, 8 to 35 percent slopes,
72E
dissected
Deerton-Tokiahok-Trout Bay complex, 15 to 70 percent slopes,
72F
dissected

Habitat
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4

76C

Garlic-Blue Lake-Voelker complex, 1 to 12 percent slopes, dissected

4

76E

Garlic-Blue Lake-Voelker complex, 8 to 35 percent slopes, dissected

4

77B
77D
77E
95B

Garlic-Blue Lake-Voelker complex, 15 to 60 percent slopes,
dissected
Garlic-Blue Lake-Voelker complex, 1 to 6 percent slopes
Garlic-Blue Lake-Voelker complex, 6 to 15 percent slopes
Garlic-Blue Lake-Voelker complex, 15 to 35 percent slopes
Liminga fine sand, 0 to 6 percent slopes

104C

Fence very fine sandy loam, 1 to 12 percent slopes, dissected

4

125B
125D
125E

Stutts-Kalkaska complex, 0 to 6 percent slopes
Stutts-Kalkaska complex, 6 to 15 percent slopes
Stutts-Kalkaska complex, 15 to 35 percent slopes
Munising, calcareous substratum-Ensley complex, 0 to 6 percent
slopes
Munising-Yalmer complex, 1 to 12 percent slopes, dissected, very
stony
Munising-Skanee complex, 0 to 6 percent slopes, stony
Shoepac-Ensley complex, 0 to 6 percent slopes
Zeba-Jacobsville complex, 0 to 3 percent slopes, very stony

4
4
4

76F

135B
145C
146B
148B
155A

272

4
4
4
4
4

4
4
4
4
4

Table 62: (cont'd)
Numeric Name
157B
Reade-Nahma complex, 0 to 6 percent slopes, stony
Munising-Abbaye fine sandy loams, 1 to 12 percent slopes,
158C
dissected, stony
160B
Paquin-Finch sands, 0 to 6 percent slopes
161B
Yellowdog-Buckroe complex, 0 to 6 percent slopes, stony
165B
Chocolay-Waiska complex, 1 to 6 percent slopes, very stony
Chocolay very stony fine sandy loam, 1 to 6 percent slopes, very
170B
stony
171B
172D
172F

Paavola very gravelly loamy sand, 0 to 6 percent slopes, very stony
Buckroe-Rock outcrop complex, 6 to 25 percent slopes, very
bouldery
Buckroe-Rock outcrop complex, 25 to 70 percent slopes, very
bouldery

Habitat
4
4
4
4
4
4
4
4
4

181E

Frohling-Tokiahok complex, 8 to 35 percent slopes, dissected, stony

4

185B
186B
186D
187B
188B
188D
188E
197B
198B

McMaster cobbly sandy loam, 0 to 4 percent slopes
Chatham fine sandy loam, 1 to 6 percent slopes, stony
Chatham fine sandy loam, 6 to 15 percent slopes, stony
Reade silt loam, 0 to 4 percent slopes
Eben very cobbly sandy loam, 1 to 6 percent slopes, stony
Eben very cobbly sandy loam, 6 to 15 percent slopes, stony
Eben very cobbly sandy loam, 15 to 35 percent slopes, stony
Shoepac-Trenary silt loams, 1 to 6 percent slopes
Shoepac-Reade silt loams, 1 to 4 percent slopes

4
4
4
4
4
4
4
4
4

202B

Sauxhead sandy loam, 1 to 6 percent slopes, rocky, very stony

4

206B
206D
211B
214B
214D
214E
225B
225D
226B
226D
226E
226F
227A
232B
233B

Traunik cobbly fine sandy loam, 1 to 6 percent slopes
Traunik cobbly fine sandy loam, 6 to 15 percent slopes
Munising-Abbaye fine sandy loams, 1 to 6 percent slopes
Kalkaska-Blue Lake complex, 1 to 6 percent slopes
Kalkaska-Blue Lake complex, 6 to 15 percent slopes
Kalkaska-Blue Lake complex, 15 to 35 percent slopes
Cusino loamy sand, 1 to 6 percent slopes
Cusino loamy sand, 6 to 15 percent slopes
Kalkaska-Cusino complex, 1 to 6 percent slopes
Kalkaska-Cusino complex, 6 to 15 percent slopes
Kalkaska-Cusino complex, 15 to 35 percent slopes
Kalkaska-Cusino complex, 35 to 70 percent slopes
Halfaday sand, 0 to 3 percent slopes
Shelldrake sand, 0 to 8 percent slopes
Abbaye-Zeba complex, 0 to 6 percent slopes, very stony

4
4
4
4
4
4
4
4
4
4
4
4
4
4
4

273

Table 62: (cont'd)
Numeric Name

Habitat

235B

Sauxhead-Burt complex, 0 to 4 percent slopes, rocky, very stony

4

236B

Waiska stony sandy loam, 1 to 6 percent slopes, extremely bouldery

4

251B
251D

Waiska stony sandy loam, 6 to 15 percent slopes, extremely
bouldery
Chatham-Davies complex, 0 to 6 percent slopes
Longrie-Shingleton complex, 1 to 6 percent slopes
Trout Bay-Gongeau-Shingleton-Rock outcrop complex, 1 to 70
percent slopes
Garlic sand, 0 to 6 percent slopes
Garlic sand, 6 to 15 percent slopes
Garlic sand, 15 to 35 percent slopes
Escanaba-Greylock complex, 1 to 6 percent slopes
Escanaba-Greylock complex, 6 to 15 percent slopes
Escanaba-Greylock complex, 15 to 35 percent slopes
Chocolay-Jacobsville complex, 0 to 6 percent slopes, extremely
stony
Greylock fine sandy loam, 1 to 6 percent slopes
Greylock fine sandy loam, 6 to 15 percent slopes

254C

Kalkaska-Blue Lake complex, 1 to 12 percent slopes, dissected

4

254E

Kalkaska-Blue Lake complex, 8 to 35 percent slopes, dissected

4

254F

Kalkaska-Blue Lake complex, 15 to 70 percent slopes, dissected

4

256B
268C

4
4

281E
282B
282D
284B

Whitewash sand, 0 to 4 percent slopes
Munising, calcareous substratum-Frohling, 1 to 12
Frohling, calcareous substratum-Garlic-Cookson complex, 8 to 35
percent slopes, dissected
Munising-Yalmer-Frohling complex, calcareous substratum, 1 to 12
percent slopes, dissected
Munising, calcareous substratum-Cookson fine sandy loams, 1 to 6
percent slopes
Mongo silt loam, 8 to 45 percent slopes, dissected
Furlong-Shingleton complex, 1 to 6 percent slopes
Furlong-Shingleton complex, 6 to 15 percent slopes
Steuben-Blue Lake-Kalkaska complex, 1 to 6 percent slopes

284D

Steuben-Blue Lake-Kalkaska complex, 6 to 15 percent slopes

4

284E

Steuben-Blue Lake-Kalkaska complex, 15 to 35 percent slopes

4

236D
237B
239B
240F
246B
246D
246E
248B
248D
248E
250B

269E
272C
275B

274

4
4
4
4
4
4
4
4
4
4
4
4
4

4
4
4
4
4
4
4

Table 62:
Numeric
285B
286B
287B

(cont'd)
Name
Halfaday-Kinross complex, 0 to 4 percent slopes
Greylock-Cookson fine sandy loams, 1 to 6 percent slopes
McMaster-Davies complex, 0 to 4 percent slopes

Habitat
4
4
4

292B

Mashek fine sandy loam, sandy substratum, 0 to 4 percent slopes

4

296D
296E
299F

Island Lake-McMillan complex, 6 to 15 percent slopes
Island Lake-McMillan complex, 15 to 35 percent slopes
Shelldrake fine sand, 2 to 75 percent slopes

4
4
4

301F

Cookson-Nykanen complex, 15 to 50 percent slopes, dissected

4

302B
302D
302E
302F
303B
303D
303E

Dillingham-Kalkaska complex, 1 to 6 percent slopes
Dillingham-Kalkaska complex, 6 to 15 percent slopes
Dillingham-Kalkaska complex, 15 to 35 percent slopes
Dillingham-Kalkaska complex, 35 to 70 percent slopes
Kiva-Trenary fine sandy loams, 1 to 6 percent slopes
Kiva-Trenary fine sandy loams, 6 to 15 percent slopes
Kiva-Trenary fine sandy loams, 15 to 35 percent slopes

4
4
4
4
4
4
4

306C

Deerton-Tokiahok-Jeske complex, 1 to 12 percent slopes, dissected

4

311B

Kalkaska sand, 0 to 6 percent slopes, very deep water table, burned

4

312B
312D

Kalkaska sand, 6 to 15 percent slopes, very deep water table,
burned
Island Lake sand, 0 to 6 percent slopes, burned
Island Lake sand, 6 to 15 percent slopes, burned

313B

Kalkaska sand, 0 to 6 percent slopes, deep water table, burned

311D

4
4
4
4

316B
316D
317B

Blue Lake loamy sand, 0 to 6 percent slopes, very deep water table,
burned
Blue Lake loamy sand, 0 to 6 percent slopes, deep water table,
burned
Blue Lake loamy sand, 0 to 6 percent slopes, burned
Blue Lake loamy sand, 6 to 15 percent slopes, burned
Kalkaska sand, 0 to 6 percent slopes, very deep water table

317D

Kalkaska sand, 6 to 15 percent slopes, very deep water table

4

318B

Island Lake sand, 0 to 6 percent slopes, very deep water table

4

318D

Island Lake sand, 6 to 15 percent slopes, very deep water table

4

319B
319D

Island Lake sand, 0 to 6 percent slopes
Island Lake sand, 6 to 15 percent slopes

4
4

314B
315B

275

4
4
4
4
4

Table 62:
Numeric
319E
319F
320B
321B
321D
17A
18
19
21A
38B
46
48
51
57
58
59

88
93
147A
166
167
191B
200A
221B
234A
241
243
245B
249B
252A
266A
267A
33
42

(cont'd)
Name
Island Lake sand, 15 to 35 percent slopes
Island Lake sand, 35 to 60 percent slopes
Kalkaska sand, 0 to 6 percent slopes, deep water table
Kalkaska-Deerton sands, 0 to 6 percent slopes
Kalkaska-Deerton sands, 6 to 15 percent slopes
Au Gres sand, 0 to 3 percent slopes
Kinross muck
Deford muck
Ingalls sand, 0 to 3 percent slopes
Rhody-Towes complex, 0 to 4 percent slopes
Jacobsville muck, very stony
Burt muck
Nahma-Ruse complex
Carbondale, Lupton, and Tawas soils
Dawson, Greenwood, and Loxley soils
Chippeny-Nahma mucks
Jeske-Gongeau-Deerton complex, bedrock terrace, 1 to 20 percent
slopes
Jeske-Gongeau-Deerton complex, bedrock terrace, 1 to 45 percent
slopes
Cathro-Ensley mucks
Tawas-Deford mucks
Skanee-Gay complex, 0 to 3 percent slopes, very stony
Skandia mucky peat
Skandia-Jacobsville complex, stony
Ruse-Ensign complex, 0 to 3 percent slopes
Charlevoix-Ensley complex, 0 to 3 percent slopes
Jeske-Au Train-Gongeau complex, 0 to 8 percent slopes
Levasseur-Burt complex, 0 to 3 percent slopes, very stony
Cathro-Gay mucks
Markey mucky peat
Trout Bay-Lupton-Gongeau complex, 0 to 6 percent slopes
Sauxhead-Skandia complex, 0 to 4 percent slopes
Finch-Kinross complex, 0 to 3 percent slopes
Spot-Finch complex, 0 to 3 percent slopes
Finch sand, 0 to 3 percent slopes
Ensley muck
Davies very cobbly muck

71A

Evart-Sturgeon silt loams, 0 to 2 percent slopes, frequently flooded

6

60
W

Histosols and Aquents, ponded
Water

7
7

65D
65F

276

Habitat
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
6
6

Numeric
33
50
116
119
122
143
144
153

Name
Pits, sand and gravel
Deford fine sand
Udorthents, nearly level
Gogomain very fine sandy loam
Pits, quarry
Burleigh loamy fine sand
Urban land-Udorthents complex, nearly level
Dumps, limestone

106A

Potagannissing-Rock outcrop complex, 0 to 3 percent slopes

0

108D
117B
117D
117F
121B
123B
124D
136A
147B
147D
156A
29A
52A
53B
56A

Shelter-Alpena complex, 0 to 15 percent slopes
Manistee sand, 0 to 6 percent slopes
Manistee sand, 6 to 15 percent slopes
Manistee sand, 25 to 50 percent slopes
Rockbottom stony silt loam, 2 to 6 percent slopes
Ocqueoc fine sand, 0 to 6 percent slopes
Alpena very cobbly sandy loam, 0 to 15 percent slopes
Westbury-Gay complex, 0 to 3 percent slopes
Shelter very stony loam, 0 to 6 percent slopes
Shelter very stony loam, 6 to 15 percent slopes
Rockcut-Pinconning complex, 0 to 3 percent slopes
Solona fine sandy loam, 0 to 3 percent slopes
Ingalls loamy sand, 0 to 3 percent slopes
Menominee loamy sand, 0 to 6 percent slopes
Ensign silt loam, 0 to 3 percent slopes, rocky

5
4
4
4
4
4
5
5
5
5
5
5
6
4
5

57B

Summerville-Longrie complex, 1 to 6 percent slopes, rocky

4

78B
Waiska sandy loam, 0 to 6 percent slopes
86A
Ingalls-Halfaday complex, 0 to 3 percent slopes
97A
Wega very fine sandy loam, 0 to 3 percent slopes
99A
Westbury stony fine sandy loam, 0 to 3 percent slopes
17D
Deer Park fine sand, 0 to 15 percent slopes
17F
Deer Park fine sand, 25 to 50 percent slopes
38F
Deer Park-Kinross complex, 0 to 50 percent slopes
18B
Rubicon sand, 0 to 6 percent slopes
18D
Rubicon sand, 6 to 15 percent slopes
18E
Rubicon sand, 15 to 35 percent slopes
98
Ermatinger silt loam
126
Pickford silt loam
103D
Velvet-Rockbottom complex, 6 to 15 percent slopes
103E
Velvet-Rockbottom complex, 15 to 35 percent slopes
10B
Ontonagon silt loam, 2 to 6 percent slopes
Table 63: Chippewa County

277

Habitat
0
5
0
5
0
5
0
0

4
6
4
4
1
1
1
2
2
2
3
3
3
3
3

Table 63: (cont'd)
Numeric
Name
10D
Ontonagon silt loam, 6 to 15 percent slopes
10F
Ontonagon silt loam, 25 to 50 percent slopes
114B
Velvet very stony loamy sand, 0 to 6 percent slopes
114D
Velvet very stony loamy sand 6 to 15 percent slopes
125B
Croswell-Markey complex, 0 to 6 percent slopes
139A
Rudyard-Urban land complex, 0 to 3 percent slopes
20A
Croswell sand, 0 to 3 percent slopes

Habitat
3
3
3
3
3
3
3

41D

Summerville-Rock outcrop complex, 1 to 15 percent slopes

3

41F

Summerville-Rock outcrop complex, 15 to 45 percent slopes

3

67B

Duel-Rock outcrop complex, 1 to 6 percent slopes

3

87B

Rousseau fine sand, moderately wet, 0 to 6 percent slopes

3

88A
91B
91D
91E
93F
96B
104B
104D
104F
107B
132B
132F
135B

3
3
3
3
3
3
4
4
4
4
4
4
4

13B
13D
13F
148B
149B

Croswell-Au Gres sands, 0 to 3 percent slopes
Rousseau fine sand, 0 to 6 percent slopes
Rousseau fine sand, 6 to 15 percent slopes
Rousseau fine sand, 15 to 35 percent slopes
Ontonagon-Pickford complex, 0 to 50 percent slopes
Velvet-Westbury complex, 0 to 6 percent slopes
Amasa very fine sandy loam, 0 to 6 percent slopes
Amasa very fine sandy loam, 6 to 15 percent slopes
Amasa very fine sandy loam, 25 to 50 percent slopes
Oldman stony fine sandy loam, 2 to 6 percent slopes
Sugar very fine sandy loam, 0 to 6 percent slopes
Sugar very fine sandy loam, 25 to 50 percent slopes
Longrie-Posen complex, 0 to 6 percent slopes
Rousseau, dark subsoil-Urban land complex, 0 to 4 percent
slopes
Alcona loamy very fine sand, 0 to 6 percent slopes
Alcona loamy very fine sand, 6 to 15 percent slopes
Alcona loamy very fine sand, 25 to 50 percent slopes
Longrie-Rock outcrop complex, 1 to 6 percent slopes
Kalkaska sand, 0 to 6 percent slopes, stony

159B

Amasa-Sugar very fine sandy loams, 0 to 6 percent slopes

4

159F

Amasa-Sugar very fine sandy loam, 25 to 50 percent slopes

4

15B

Rousseau fine sand, dark subsoil, 0 to 6 percent slopes

4

15D

Rousseau fine sand, dark subsoil, 6 to 15 percent slopes

4

138B

278

4
4
4
4
4
4

Table 63: (cont'd)
Numeric
Name

Habitat

15E

Rousseau fine sand, dark subsoil, 15 to 35 percent slopes

4

15F

Rousseau fine sand, dark subsoil, 35 to 60 percent slopes

4

19B
19D
19E
19F
25B
25D
27B
28B
42D
44B
44D
44E
46B
46D
46E
61A
79B
79D
80B

Kalkaska sand, 0 to 6 percent slopes
Kalkaska sand, 6 to 15 percent slopes
Kalkaska sand, 15 to 35 percent slopes
Kalkaska sand, 35 to 60 percent slopes
Guardlake loam, 0 to 6 percent slopes
Guardlake loam, 6 to 15 percent slopes
Emmet sandy loam, 1 to 6 percent slopes
Longrie sandy loam, 1 to 6 percent slopes
Emmet-Kalkaska complex, 1 to 15 percent slopes
Posen stony fine sandy loam, 1 to 6 percent slopes
Posen stony fine sandy loam, 6 to 15 percent slopes
Poson stony fine sandy loam, 15 to 35 percent slopes
Pence loamy sand, 0 to 6 percent slopes
Pence loamy sand, 6 to 15 percent slopes
Pence loamy sand, 15 to 35 percent slopes
Halfaday sand, 0 to 3 percent slopes
Kalkaska-Manistee sands, 0 to 6 percent slopes
Kalkaska-Manistee sands, 6 to 15 percent slopes
Superior fine sandy loam, 1 to 6 percent slopes

4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4

84B

Rousseau, dark subsoil-Alcona complex, 0 to 6 percent slopes

4

84D

Rousseau, dark subsoil-Alcona complex, 6 to 15 percent slopes

4

84F
85B
12
22
23
36
37
68
101
102
111
112
113

Rousseau, dark subsoil-Alcona complex, 25 to 50 percent
slopes
Kalkaska-Ocqueoc complex, 0 to 6 percent slopes
Pickford silty clay loam
Kinross muck
Roscommon muck
Markey and Carbondale mucks
Dawson and Loxley peats
Pinconning mucky loamy sand
Chippeny muck
Kinross-Dawson complex
Gutport muck
Soo silty clay loam
Ruse mucky fine sandy loam

279

4
4
5
5
5
5
5
5
5
5
5
5
5

Table 63: (cont'd)
Numeric
Name
127
Gay stony muck
133
Dora muck
150
Fibre muck
151
Beavertail muck
152
Grousehaven muck
11A
Rudyard silty clay loam, 0 to 3 percent slopes
128F
Alcona-Markey complex, 0 to 50 percent slopes
129A
Rudyard silt loam, 0 to 3 percent slopes
130A
Rudyard-Pickford silty clay loams, 0 to 3 percent slopes
137A
Kinross-Wainola complex, 0 to 3 percent slopes

Habitat
5
5
5
5
5
5
5
5
5
5

145A

Gaastra-Gogomain-Ingalls complex, 0 to 3 percent slopes

5

146A
14A

Allendale-Fibre complex, 0 to 3 percent slopes
Gaastra silt loam, 0 to 3 percent slopes

5
5

154F

Dawson-Deer Park-Wainola complex, 0 to 50 percent slopes

5

155B

Allendale-Posen-Pickford complex, 0 to 6 percent slopes

5

21A
32A

Au Gres sand, 0 to 3 percent slopes
Allendale loamy fine sand, 0 to 3 percent slopes

5
5

39D

Au Gres-Dawson-Rubicon complex, 0 to 15 percent slopes

5

40A

Rudyard-Allendale complex, 0 to 3 percent slopes

5

48E

Wainola-Kinross-Rousseau complex, 0 to 35 percent slopes

5

49A
83A
89A
92A

Wainola fine sand, 0 to 3 percent slopes
Allendale-Croswell complex, 0 to 3 percent slopes
Kinross-Au Gres complex, 0 to 3 percent slopes
Biscuit very fine sandy loam, 0 to 3 percent slopes

5
5
5
5

94A

Markey-Kinross-Au Gres complex, 0 to 3 percent slopes

5

95A
34
35
W

Bowers silt loam, 0 to 3 percent slopes
Entisols, flooded
Histosols and Aquents, ponded
Water

5
7
7
7

280

Numeric
Bp
BtA
FaA
FaB
Gw
KlA
KnB
KnD
Lm
Ma
MlB
NsA
NsB
OtB
PfA
PkA
Pq
RkB
Sl
YaB
YaD
GrB
GrD
EdB
EeB
RuB
RuD
RuE
KdB
KdD
RoB
RoD
RsD
AlC
BlB
BlD
BlE
BoB
BoD
BrA
CrA
DuB
Table 64:

Name
Borrow pits
Brimley fine sandy loam, 0 to 4 percent slopes
Fairport silt loam, 0 to 2 percent slopes
Fairport silt loam, 2 to 6 percent slopes
Greenwood peat
Kawkawlin silt loam, 0 to 2 percent slopes
Keweenaw loamy sand, 0 to 6 percent slopes
Keweenaw loamy sand, 6 to 18 percent slopes
Limestone rock land
Made land
Melita sand, 0 to 6 percent slopes
Nester silt loam, 0 to 2 percent slopes
Nester silt loam, 2 to 6 percent slopes
Otisco loamy sand, 0 to 6 percent slopes
Algonquin silt loam, 0 to 4 percent slopes
Algonquin-Pickford complex, 0 to 4 percent slopes
Pits, quarry
Roscommon-Kalkaska sands, 0 to 6 percent slopes
Sewage lagoons
Yalmer sand, 0 to 6 percent slopes
Yalmer sand, 6 to 18 percent slopes
Grayling sand, 0 to 6 percent slopes
Grayling sand, 6 to 18 percent slopes
Eastport sand, 0 to 6 percent slopes
Eastport-Roscommon sands, 0 to 6 percent slopes
Rubicon sand, 0 to 6 percent slopes
Rubicon sand, 6 to 18 percent slopes
Rubicon sand, 18 to 40 percent slopes
Karlin sandy loam, 0 to 6 percent slopes
Karlin sandy loam, 6 to 18 percent slopes
Rousseau fine sand, 0 to 6 percent slopes
Rousseau fine sand, 6 to 18 percent slopes
Rousseau fine sand, hilly
Alpena gravelly sandy loam, 0 to 12 percent slopes
Blue Lake sand, 0 to 6 percent slopes
Blue Lake sand, 6 to 18 percent slopes
Blue Lake sand, 18 to 40 percent slopes
Bohemian fine sandy loam, 0 to 6 percent slopes
Bohemian fine sandy loam, 6 to 18 percent slopes
Bowers silt loam, 0 to 4 percent slopes
Croswell sand, 0 to 4 percent slopes
Duel loamy sand, 0 to 6 percent slopes
Delta County

281

Habitat
0
5
4
4
5
4
4
4
0
0
4
4
4
5
0
0
0
4
0
4
4
1
1
2
2
2
2
2
3
3
3
3
3
4
4
4
4
4
4
4
4
4

Table 64:
Numeric
EaB
EmA
EmB
EmC
EnA
GcB
KaB
KaD
KaE
KsB
KsD
LoA
LoB
LsD
MnB
MnD
OnA
OnB
OnC
OnD
SuA
TrA
TrB
TrC
TrD
WlB
WlD
AuB
AvA
Bs
Bu
Cb
Ch
Ck
ClA
Cn
Da
Dd
IoB
Kr
McB
McD

(cont'd)
Name
Springlake sand, 0 to 6 percent slopes
Emmet sandy loam, 0 to 2 percent slopes
Emmet sandy loam, 2 to 6 percent slopes
Emmet sandy loam, 6 to 12 percent slopes
Ensign fine sandy loam, 0 to 3 percent slopes
Gilchrist sand, 0 to 6 percent slopes
Kalkaska sand, 0 to 6 percent slopes
Kalkaska sand, 6 to 18 percent slopes
Kalkaska sand, 18 to 40 percent slopes
Kiva sandy loam, 0 to 6 percent slopes
Kiva sandy loam, 6 to 20 percent slopes
Longrie sandy loam, 0 to 2 percent slopes
Longrie sandy loam, 2 to 6 percent slopes
Longrie and Summerville sandy loams, 6 to 18 percent slopes
Menominee loamy sand, 0 to 6 percent slopes
Menominee loamy sand, 6 to 18 percent slopes
Onaway fine sandy loam, 0 to 2 percent slopes
Onaway fine sandy loam, 2 to 6 percent slopes
Onaway fine sandy loam, 6 to 12 percent slopes
Onaway fine sandy loam, 12 to 18 percent slopes
Summerville fine sandy loam, 0 to 4 percent slopes
Trenary fine sandy loam, 0 to 2 percent slopes
Trenary fine sandy loam, 2 to 6 percent slopes
Trenary fine sandy loam, 6 to 12 percent slopes
Trenary fine sandy loam, 12 to 18 percent slopes
Wallace sand, 0 to 6 percent slopes
Wallace sand, 6 to 18 percent slopes
Au Gres sand, 0 to 6 percent slopes
Battlefield loamy sand, 0 to 4 percent slopes
Brevort mucky loamy sand
Bruce mucky fine sandy loam, coarse variant
Carbondale, Lupton, and Rifle soils
Cathro muck
Cathro and Tacoosh mucks
Charlevoix sandy loam, 0 to 4 percent slopes
Chippeny muck
Dawson peat
Dawson and Greenwood peats
Iosco sand, 0 to 6 percent slopes
Kinross mucky sand
Mancelona loamy sand, 0 to 6 percent slopes
Mancelona loamy sand, 6 to 18 percent slopes

282

Habitat
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5

Table 64:
Numeric
Nh
Pc
Rc
Rv
ScA
SvA
SwA
Ta
WaA
Wm
Dm
Es
Ad
Mh
W

(cont'd)
Name
Nahma muck
Pickford silt loam
Roscommon mucky sand
Ruse silt loam
Finch sand, 0 to 3 percent slopes
Sundell fine sandy loam, 0 to 4 percent slopes
Sundell loamy fine sand, sandy variant, 0 to 4 percent slopes
Tawas muck
Wainola fine sand, 0 to 4 percent slopes
Wheatley mucky loamy sand
Deford loamy fine sand
Ensley and Angelica soils
Alluvial land
Marsh
Water

283

Habitat
5
5
5
5
5
5
5
5
5
5
6
6
7
7
7

Numeric
33
116
122
300
17C
17E
17F
18B
18D
18E
18F
20B
45D
45E
65B
65D
65E
88B
90D
90E
90F
91D
91E
91F
109D
109F
174B
198B
198D

Name
Pits, sand and gravel
Udipsamments and Udorthents, nearly level
Pits, quarry
Beaches
Deer Park sand, 0 to 10 percent slopes
Deer Park sand, 10 to 25 percent slopes
Deer Park sand, 25 to 60 percent slopes
Rubicon sand, 0 to 6 percent slopes
Rubicon sand, 6 to 15 percent slopes
Rubicon sand, 15 to 35 percent slopes
Rubicon sand, 35 to 60 percent slopes
Croswell sand, 0 to 6 percent slopes
Rubicon-Spot complex, 0 to 15 percent slopes
Rubicon-Spot complex, 0 to 35 percent slopes
Rubicon sand, organic surface, 0 to 6 percent slopes
Rubicon sand, organic surface, 6 to 15 percent slopes
Rubicon sand, organic surface, 15 to 35 percent slopes
Croswell-Au Gres sands, 0 to 6 percent slopes
Rousseau-Spot complex, 0 to 15 percent slopes
Rousseau-Spot complex, 0 to 35 percent slopes
Rousseau-Spot complex, 0 to 60 percent slopes
Rousseau fine sand, 6 to 15 percent slopes
Rousseau fine sand, 15 to 35 percent slopes
Rousseau fine sand, 35 to 60 percent slopes
Rousseau-Dawson complex, 0 to 15 percent slopes
Rousseau-Dawson complex, 0 to 60 percent slopes
Croswell-Spot complex, 0 to 6 percent slopes
Vilas loamy sand, 0 to 6 percent slopes
Vilas loamy sand, 6 to 15 percent slopes
Croswell, rarely flooded-Deford, frequently flooded complex, 0 to 6
201B
percent slopes
214D
Rousseau-Markey complex, 0 to 15 percent slopes
214E
Rousseau-Markey complex, 0 to 35 percent slopes
129A
Rudyard silt loam, 0 to 3 percent slopes
130A
Rudyard-Pickford silt loams, 0 to 3 percent slopes
205B
Kalkaska sand, 0 to 6 percent slopes, burned
205D
Kalkaska sand, 6 to 15 percent slopes, burned
10D
Ontonagon silt loam, 6 to 15 percent slopes
15B
Liminga fine sand, 0 to 6 percent slopes
15D
Liminga fine sand, 6 to 15 percent slopes
15E
Liminga fine sand, 15 to 35 percent slopes
15F
Liminga fine sand, 35 to 60 percent slopes
Table 65: Luce County

284

Habitat
0
0
0
0
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
4
4
4
4
4

Table 65:
Numeric
16B
19B
19D
19E
19F
24B
29A
31B
31D
31E
31F
46B
46D
46E
46F
47B
47D
53B
57B
57D
57E
61B
66B
66D
66E
66F
74B
75D
75E
75F
76D
76E
84B
84D
84E
85B
85D
85E

(cont'd)
Name
Graveraet fine sandy loam, 1 to 4 percent slopes
Kalkaska sand, 0 to 6 percent slopes
Kalkaska sand, 6 to 15 percent slopes
Kalkaska sand, 15 to 35 percent slopes
Kalkaska sand, 35 to 60 percent slopes
Springlake loamy coarse sand, 0 to 6 percent slopes
Solona fine sandy loam, 0 to 3 percent slopes
McMillan fine sandy loam, 0 to 6 percent slopes
McMillan fine sandy loam, 6 to 15 percent slopes
McMillan fine sandy loam, 15 to 35 percent slopes
McMillan fine sandy loam, 35 to 60 percent slopes
Kalkaska loamy sand, 0 to 6 percent slopes
Kalkaska loamy sand, 6 to 15 percent slopes
Kalkaska loamy sand, 15 to 35 percent slopes
Kalkaska loamy sand, 35 to 60 percent slopes
Trenary fine sandy loam, 2 to 6 percent slopes
Trenary fine sandy loam, 6 to 15 percent slopes
Menominee sand, sandy substratum, 2 to 6 percent slopes
Amadon-Longrie sandy loams, 1 to 6 percent slopes, rocky
Amadon-Longrie sandy loams, 6 to 15 percent slopes, rocky
Amadon-Longrie sandy loams, 15 to 35 percent slopes, rocky
Paquin sand, 0 to 6 percent slopes
Kalkaska-Kaks complex, 0 to 6 percent slopes
Kalkaska-Kaks complex, 6 to 15 percent slopes
Kalkaska-Kaks complex, 15 to 35 percent slopes
Kalkaska-Kaks complex, 35 to 60 percent slopes
Menominee, sandy substratum-Graveraet complex, 1 to 6 percent
slopes
Dillingham-Kalkaska complex, 6 to 15 percent slopes
Dillingham-Kalkaska complex, 15 to 35 percent slopes
Dillingham-Kalkaska complex, 35 to 70 percent slopes
Menominee, sandy substratum-Trenary complex, 6 to 15 percent
slopes
Menominee, sandy substratum-Trenary complex, 15 to 35 percent
slopes
Liminga-Alcona complex, 0 to 6 percent slopes
Liminga-Alcona complex, 6 to 15 percent slopes
Liminga-Alcona complex, 15 to 35 percent slopes
Kalkaska-Okeefe sands, 0 to 6 percent slopes
Kalkaska-Okeefe sands, 6 to 15 percent slopes
Kalkaska-Okeefe sands, 15 to 35 percent slopes

285

Habitat
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4

Table 65: (cont'd)
Numeric Name
Ontonagon-Pickford, occasionally flooded complex, 0 to 50 percent
93F
slopes
104B
Pence very fine sandy loam, 0 to 6 percent slopes
104D
Pence very fine sandy loam, 6 to 15 percent slopes
104E
Pence very fine sandy loam, 15 to 35 percent slopes
117D
Manistee sand, sandy substratum, 6 to 15 percent slopes
120B
McMillan-Trenary fine sandy loams, 6 to 15 percent slopes
120D
McMillan-Trenary fine sandy loams, 6 to 15 percent slopes
120E
McMillan-Trenary fine sandy loams, 15 to 35 percent slopes
132B
Sugar very fine sandy loam, 0 to 6 percent slopes
167D
Battydoe, stony-Wallace complex, 6 to 15 percent slopes
173B
Paquin-Finch sands, 0 to 6 percent slopes
175D
Wallace-Spot complex, 0 to 15 percent slopes
175E
Wallace-Spot complex, 0 to 35 percent slopes
176B
Paquin-Spot complex, 0 to 6 percent slopes
179B
Wallace sand, 0 to 6 percent slopes
179D
Wallace sand, 6 to 15 percent slopes
179E
Wallace sand, 15 to 35 percent slopes
179F
Wallace sand, 35 to 60 percent slopes
180B
Millecoquins silt loam, 0 to 6 percent slopes
186D
Sporley silt loam, 6 to 15 percent slopes
186E
Sporley silt loam, 15 to 35 percent slopes
186F
Sporley silt loam, 35 to 60 percent slopes
189A
Bodi-Chesbrough silt loams, 0 to 3 percent slopes
190B
Bodi silt loam, 0 to 6 percent slopes
191D
Widgeon-Kalkaska complex, 6 to 15 percent slopes
197D
Zandi silt loam, 6 to 15 percent slopes
197E
Zandi silt loam, 15 to 35 percent slopes
200B
Pence loamy sand, 0 to 6 percent slopes
200D
Pence loamy sand, 6 to 15 percent slopes
200E
Pence loamy sand, 15 to 35 percent slopes
202B
Whitewash sand, 0 to 4 percent slopes
203D
Frohling loamy sand, 8 to 15 percent slopes
203E
Frohling loamy sand, 15 to 35 percent slopes
206B
Deerton loamy sand, 0 to 6 percent slopes
211D
Frohling-Wallace complex, 6 to 15 percent slopes
211E
Frohling-Wallace complex, 15 to 35 percent slopes
215B
Wallace-Alcona complex, 0 to 6 percent slopes
215D
Wallace-Alcona complex, 6 to 15 percent slopes
246B
Garlic sand, 0 to 6 percent slopes
246D
Garlic sand, 6 to 15 percent slopes
286B
Fence silt loam, 0 to 6 percent slopes

286

Habitat
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4

Table 65:
Numeric
287B
21A
22
30
32A
36
37
60A
89A
94A
102
110D
110E
126
133
143
146A
187B
193A
194A
195A
199B
188
204
212
23
35
W

(cont'd)
Name
Noseum fine sandy loam, 0 to 4 percent slopes
Finch sand, 0 to 3 percent slopes
Spot peat
Kinross muck
Allendale loamy fine sand, 0 to 3 percent slopes
Carbondale, Lupton, and Tawas soils
Dawson, Greenwood, and Loxley soils
Kinross-Au Gres complex, 0 to 3 percent slopes
Spot-Finch complex, 0 to 3 percent slopes
Tawas-Spot-Finch complex, 0 to 3 percent slopes
Spot-Dawson peats
Au Gres-Dawson-Rubicon complex, 0 to 15 percent slopes
Au Gres-Dawson-Rubicon complex, 0 to 35 percent slopes
Pickford silt loam
Dorval muck
Caffey muck
Allendale-Fibre complex, 0 to 3 percent slopes
Auger silt loam, 0 to 6 percent slopes
Annanias silt loam, 0 to 3 percent slopes
Hendrie-Annanias complex, 0 to 3 percent slopes
Chesbrough silt loam, 0 to 3 percent slopes
Auger-Annanias silt loams, 0 to 6 percent slopes
Hendrie mucky peat
Gogomain muck
Markey mucky peat
Leafriver mucky peat
Histosols and Aquents, ponded
Water

287

Habitat
4
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
6
7
7

Numeric
122
33
70F
116
124D
20B
125B
88B
17D
17E
38E
170B
170E
170D
18B
18E
18F
18D
57B
57D
41D
41F
95A
185
10F
10D

Name
Pits, quarry
Pits, sand and gravel
St. Ignace-Rock outcrop complex, 35 to 70 percent slopes
Udipsamments and Udorthents, nearly level
Alpena gravelly loam, 0 to 15 percent slopes
Croswell sand, 0 to 6 percent slopes
Croswell-Markey complex, 0 to 6 percent slopes
Croswell-Wainola complex, 0 to 6 percent slopes
Eastport sand, 0 to 15 percent slopes
Eastport sand, 15 to 35 percent slopes
Eastport-Leafriver complex, 0 to 35 percent slopes
Pullup fine sand, 0 to 6 percent slopes
Pullup fine sand, 15 to 35 percent slopes
Pullup fine sand, 6 to 15 percent slopes
Rubicon sand, 0 to 6 percent slopes
Rubicon sand, 15 to 35 percent slopes
Rubicon sand, 35 to 60 percent slopes
Rubicon sand, 6 to 15 percent slopes
Amadon-Longrie sandy loams, 1 to 6 percent slopes, rocky
Amadon-Longrie sandy loams, 6 to 15 percent slopes, rocky
Amadon-Rock outcrop complex, 1 to 15 percent slopes
Amadon-Rock outcrop complex, 15 to 45 percent slopes
Bowers silt loam, 0 to 3 percent slopes
Ermatinger silt loam
Ontonagon silt loam, 25 to 50 percent slopes
Ontonagon silt loam, 6 to 15 percent slopes
Ontonagon-Pickford, occasionally flooded complex, 0 to 50 percent
93F
slopes
46B
Adams sandy loam, 0 to 6 percent slopes
13B
Alcona fine sandy loam, 0 to 6 percent slopes
13D
Alcona fine sandy loam, 6 to 15 percent slopes
44B
Battydoe fine sandy loam, 1 to 6 percent slopes, stony
44E
Battydoe fine sandy loam, 15 to 35 percent slopes, stony
44D
Battydoe fine sandy loam, 6 to 15 percent slopes, stony
167B
Battydoe, stony-Wallace complex, 0 to 6 percent slopes
167E
Battydoe, stony-Wallace complex, 15 to 35 percent slopes
167D
Battydoe, stony-Wallace complex, 6 to 15 percent slopes
183B
Cozy cobbly fine sandy loam, 0 to 6 percent slopes
174B
Croswell-Spot complex, 0 to 6 percent slopes
67B
Furlong sand, 0 to 6 percent slopes, rocky
16B
Graveraet fine sandy loam, 1 to 6 percent slopes
16D
Graveraet fine sandy loam, 6 to 15 percent slopes
Table 66: Mackinac County

288

Habitat
0
0
0
0
0
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
4
4
4
4

Table 66:
Numeric
27B
27F
27D
100B
100D
25B
25E
25D
182B
19B
19E
19F
19D
28B
28D
135B
135D
61B
173B
176B
69B
24B
24E
24D
132F
132B

(cont'd)
Name
Greylock fine sandy loam, 1 to 6 percent slopes
Greylock fine sandy loam, 35 to 60 percent slopes
Greylock fine sandy loam, 6 to 15 percent slopes
Greylock-Adams complex, 0 to 6 percent slopes
Greylock-Adams complex, 6 to 15 percent slopes
Guardlake fine sandy loam, 0 to 6 percent slopes
Guardlake fine sandy loam, 15 to 35 percent slopes
Guardlake fine sandy loam, 6 to 15 percent slopes
Heinz sandy loam, 0 to 6 percent slopes
Kalkaska sand, 0 to 6 percent slopes
Kalkaska sand, 15 to 35 percent slopes
Kalkaska sand, 35 to 60 percent slopes
Kalkaska sand, 6 to 15 percent slopes
Longrie sandy loam, 1 to 6 percent slopes, rocky
Longrie sandy loam, 6 to 15 percent slopes, rocky
Longrie-Battydoe, stony complex, 1 to 6 percent slopes
Longrie-Battydoe, stony complex, 6 to 15 percent slopes
Paquin sand, 0 to 6 percent slopes
Paquin-Finch sands, 0 to 6 percent slopes
Paquin-Spot complex, 0 to 6 percent slopes
Satago silt loam, 1 to 6 percent slopes
Springlake loamy coarse sand, 0 to 6 percent slopes
Springlake loamy coarse sand, 15 to 35 percent slopes
Springlake loamy coarse sand, 6 to 15 percent slopes
Superior fine sandy loam, 25 to 50 percent slopes
Superior fine sandy loam, till substratum, 1 to 6 percent slopes

Habitat
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4

132D

Superior fine sandy loam, till substratum, 6 to 15 percent slopes

4

179B
179E
179F
179D
84B
84F
84D
175D
71B
117B
53B
53D
180B

Wallace sand, 0 to 6 percent slopes
Wallace sand, 15 to 35 percent slopes
Wallace sand, 35 to 60 percent slopes
Wallace sand, 6 to 15 percent slopes
Wallace-Alcona complex, 0 to 6 percent slopes
Wallace-Alcona complex, 35 to 60 percent slopes
Wallace-Alcona complex, 6 to 15 percent slopes
Wallace-Spot complex, 0 to 15 percent slopes
Johnswood cobbly silt loam, 2 to 6 percent slopes
Manistee sand, 0 to 6 percent slopes
Menominee loamy sand, 0 to 6 percent slopes
Menominee loamy sand, 6 to 15 percent slopes
Millecoquins very fine sandy loam, 1 to 6 percent slopes

4
4
4
4
4
4
4
4
4
4
4
4
4

289

Table 66: (cont'd)
Numeric Name

Habitat

177B

Millecoquins-Superior, till substratum complex, 1 to 6 percent slopes

4

177D

Millecoquins-Superior, till substratum complex, 6 to 15 percent slopes

4

169E
70B
70D
32A
146A
43
151
123B
143
168B
37
178B
133
21A
39E
62A
36
94A
181A
164A
12
11A
113
112
22
89A
49A
48E
68
161
92A
165A
56A
160B
163B
98
119

Ontonagon-Fluvaquents, frequently flooded complex, 0 to 35 percent
slopes
St. Ignace silt loam, 0 to 6 percent slopes
St. Ignace silt loam, 6 to 15 percent slopes, rocky
Allendale fine sand, 0 to 3 percent slopes
Allendale-Wakeley complex, 0 to 3 percent slopes
Angelica muck
Beavertail muck
Borgstrom sand, 0 to 6 percent slopes
Caffey muck
Caffey-Ingalls-Iosco complex, 0 to 6 percent slopes
Dawson and Loxley peats
Dinkey muck, 0 to 6 percent slopes
Dorval muck
Finch sand, 0 to 3 percent slopes
Finch-Dawson-Pullup complex, 0 to 35 percent slopes
Iosco sand, 0 to 3 percent slopes
Markey and Carbondale mucks
Markey-Spot-Finch complex, 0 to 3 percent slopes
Mattix sandy loam, 0 to 3 percent slopes
Moltke loam, 0 to 3 percent slopes
Pickford silty clay loam
Rudyard silty clay loam, 0 to 3 percent slopes
Ruse mucky loam
Soo silty clay loam
Spot muck
Spot-Finch complex, 0 to 3 percent slopes
Wainola fine sand, 0 to 3 percent slopes
Wainola-Leafriver-Pullup complex, 0 to 35 percent slopes
Wakeley muck
Zela muck
Engadine fine sandy loam, 0 to 3 percent slopes
Engadine-Rudyard complex, 0 to 3 percent slopes
Ensign fine sandy loam, 0 to 3 percent slopes, rocky
Esau extremely gravelly sandy loam, 0 to 3 percent slopes
Esau-Zela complex, 0 to 3 percent slopes
Glawe silt loam
Gogomain very fine sandy loam

290

4
4
4
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5

Table 66:
Numeric
166
40A
64A
147B
147D
108D
29A
23
172B
34
35
W

(cont'd)
Name
Gogomain-Pickford complex
Rudyard-Allendale complex, 0 to 3 percent slopes
Search very fine sandy loam, 0 to 3 percent slopes
Shelter very cobbly loam, 0 to 6 percent slopes, stony
Shelter very cobbly loam, 6 to 15 percent slopes, stony
Shelter-Alpena complex, 0 to 15 percent slopes, stony
Solona loam, 0 to 3 percent slopes
Leafriver mucky peat
Leafriver-Croswell-Wainola complex, 0 to 6 percent slopes
Entisols, frequently flooded
Histosols and Aquents, ponded
Water

291

Habitat
5
5
5
5
5
5
5
6
6
7
7
7

Numeric
33
76A
116F
119
218
25B
68F
70F
12B
12D
12E
12F
80F
86B
219B

Name
Pits, sand and gravel
Shuberts-Manistique complex, 0 to 3 percent slopes
Udipsamments and Udorthents, nearly level to very steep
Landfill
Pits, quarry
Proper sand, 0 to 6 percent slopes
Deer Park sand, 25 to 60 percent slopes
Deer Park-Deford-Tawas complex, 0 to 60 percent slopes
Rubicon sand, 0 to 6 percent slopes
Rubicon sand, 6 to 15 percent slopes
Rubicon sand, 15 to 35 percent slopes
Rubicon sand, 35 to 60 percent slopes
Deer Park-Dawson-Wainola complex, 0 to 60 percent slopes
Wurtsmith-Tawas-Deford complex, 0 to 6 percent slopes
Cublake sand, 0 to 6 percent slopes

491B

Neconish-Deford, rarely flooded-Wainola complex, 0 to 6 percent slopes

492D
Wurtsmith-Duck-Rubicon complex, 0 to 15 percent slopes
515B
Vilas loamy sand, 0 to 6 percent slopes
520B
Rubicon-Sultz complex, 0 to 6 percent slopes
520D
Rubicon-Sultz complex, 6 to 15 percent slopes
525B
Neconish-Kinross-Wainola complex, 0 to 6 percent slopes
547B
Rubicon sand, 0 to 6 percent slopes, very deep water table
547D
Rubicon sand, 6 to 15 percent slopes, very deep water table
565B
Rubicon sand, 0 to 6 percent slopes, deep water table
565D
Rubicon sand, 6 to 15 percent slopes, deep water table
17A
Au Gres sand, 0 to 3 percent slopes
155B
Karlin loamy fine sand, 0 to 6 percent slopes
514B
Kalkaska sand, 0 to 6 percent slopes, burned
514D
Kalkaska sand, 6 to 15 percent slopes, burned
514E
Kalkaska sand, 15 to 50 percent slopes, burned
534
Pickford silty clay loam
10B
Amadon-Rock outcrop complex, 1 to 6 percent slopes
11B
Kalkaska loamy sand, 0 to 6 percent slopes
13B
Kalkaska sand, 0 to 6 percent slopes
13D
Kalkaska sand, 6 to 15 percent slopes
13E
Kalkaska sand, 15 to 35 percent slopes
13F
Kalkaska sand, 35 to 60 percent slopes
16A
Paquin sand, 0 to 3 percent slope
20E
Rousseau-Neconish-Finch complex, 0 to 25 percent slopes
21B
Garlic sand, 0 to 6 percent slopes
21D
Garlic sand, 6 to 15 percent slopes
Table 67: Schoolcraft County

292

Habitat
0
0
0
0
0
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4

Table 67: (cont'd)
Numeric Name
21E
Garlic sand, 15 to 35 percent slopes
22E
Rousseau-Neconish-Deford complex, 0 to 25 percent slopes
24B
Rousseau fine sand, 0 to 6 percent slopes
24D
Rousseau fine sand, 6 to 15 percent slopes
24E
Rousseau fine sand, 15 to 35 percent slopes
34B
Liminga fine sand, 0 to 6 percent slopes
34E
Liminga fine sand, 15 to 35 percent slopes
56B
Shuberts-Wurtsmith-Meehan complex, 0 to 6 percent slopes
62B
McMillan-Greylock fine sandy loams, 1 to 6 percent slopes
62D
McMillan-Greylock fine sandy loams, 6 to 15 percent slopes
62E
Greylock-McMillan fine sandy loams, 15 to 35 percent slopes
Pelkie, occasionally flooded-Deford,frequently flooded, complex, 0 to 4
67B
percent slopes
73B
Graveraet-Gulliver complex, 1 to 6 percent slopes
78B
Heinz sandy loam, 0 to 6 percent slopes
87B
Longrie-Amadon silt loams, 0 to 6 percent slopes, rocky
87D
Amadon-Longrie silt loams, 6 to 15 percent slopes, rocky
87E
Amadon-Longrie-Rock outcrop complex, 15 to 35 percent slopes
88B
Cookson-Amadon silt loams, 1 to 6 percent slopes
89B
Cookson-Trenary silt loams, 1 to 6 percent slopes
98B
Guardlake fine sandy loam, 0 to 6 percent slopes
98D
Guardlake fine sandy loam, 6 to 15 percent slopes
98E
Guardlake fine sandy loam, 15 to 35 percent slopes
125B
Stutts-Kalkaska complex, 0 to 6 percent slopes
125D
Stutts-Kalkaska complex, 6 to 15 percent slopes
125E
Stutts-Kalkaska complex, 15 to 35 percent slopes
131B
Furlong-Shingleton complex, 1 to 6 percent slopes
141A
Halfaday sand, 0 to 3 percent slopes
145B
Noseum fine sandy loam, 0 to 4 percent slopes
160B
Paquin-Finch sands, 0 to 6 percent slopes
225B
Cusino loamy sand, 1 to 6 percent slopes
248B
Escanaba-Greylock complex, 1 to 6 percent slopes
287B
Hiawatha fine sandy loam, 0 to 6 percent slopes
289B
Wallace sand, 0 to 6 percent slopes
289D
Wallace sand, 6 to 15 percent slopes
289E
Wallace sand, 15 to 35 percent slopes
Munising-Yalmer-Frohling complex, calcareous substratum, 1 to 6 percent
294B
slopes
295E
Dillingham-Kalkaska complex, 15 to 35 percent slopes
355B
Springlake loamy sand, 0 to 6 percent slopes
366B
Gilchrist sand, 0 to 6 percent slopes
367B
Cozy fine sandy loam, 0 to 6 percent slopes

293

Habitat
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4

Table 67: (cont'd)
Numeric Name
400B
Amadon-Rock outcrop complex, 1 to 6 percent slopes
Pelkie, occasionally flooded-Arnheim, frequently flooded complex, 0 to 4
490B
percent slopes
505B
Graveraet fine sandy loam, 1 to 4 percent slopes
505D
Graveraet fine sandy loam, 6 to 15 percent slopes
512A
Growton fine sandy loam, 0 to 3 percent slopes
518A
Deford-Seney complex, frequently flooded, 0 to 3 percent slopes
519B
Trenary silt loam, 2 to 6 percent slopes
526B
Graveraet-Angelica complex, 0 to 4 percent slopes
527B
Islandlake-McMillan complex, 0 to 6 percent slopes
527D
Islandlake-McMillan complex, 6 to 15 percent slopes
527E
Islandlake-McMillan complex, 15 to 35 percent slopes
531B
Greylock fine sandy loam, 1 to 6 percent slopes
531D
Greylock fine sandy loam, 6 to 15 percent slopes
535B
Trenary fine sandy loam, 2 to 6 percent slopes
535D
Trenary fine sandy loam, 6 to 15 percent slopes
536B
Menominee sand, sandy substratum, 2 to 6 percent slopes
537B
McMillan fine sandy loam, 0 to 6 percent slopes
537D
McMillan fine sandy loam, 6 to 15 percent slopes

Habitat
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4

538B

Menominee, sandy substratum-McMillan complex, 0 to 6 percent slopes

4

546B
546D
548B
549B
549D
549E
551B
551D
554B
555D

Kalkaska sand, 0 to 6 percent slopes, deep water table
Kalkaska sand, 6 to 15 percent slopes, deep water table
McMillan-Trenary fine sandy loams, 1 to 6 percent slopes
Islandlake sand, 0 to 6 percent slopes
Islandlake sand, 6 to 15 percent slopes
Islandlake sand, 15 to 35 percent slopes
Kalkaska sand, 0 to 6 percent slopes, very deep water table
Kalkaska sand, 6 to 15 percent slopes, very deep water table
Duck-Halfaday complex, 0 to 6 percent slopes
Hiawatha-Deer Park-Rubicon complex, 0 to 15 percent slopes

4
4
4
4
4
4
4
4
4
4

557B

Kalkaska sand, 0 to 6 percent slopes, very deep water table, burned

4

557D

Kalkaska sand, 6 to 15 percent slopes, very deep water table, burned

4

558B
558D
559B
560B
561A
562B
563B

Islandlake sand, 0 to 6 percent slopes, burned
Islandlake sand, 6 to 15 percent slopes, burned
Kalkaska sand, 0 to 6 percent slopes, deep water table, burned
Islandlake sand, 0 to 6 percent slopes, very deep water table
Croswell sand, 0 to 3 percent slopes
Croswell-Kinross complex, 0 to 6 percent slopes
Halfaday-Kinross complex, 0 to 4 percent slopes

4
4
4
4
4
4
4

294

Table 67: (cont'd)
Numeric Name
18
Kinross muck
19A
Au Gres-Deford complex, 0 to 3 percent slopes
26A
Spot-Finch complex, 0 to 3 percent slopes
27A
Hendrie-Annanias complex, 0 to 3 percent slopes
30
Hendrie mucky peat
36
Carbondale, Lupton, and Tawas soils
37
Dawson, Greenwood, and Loxley soils
61A
Ingalls-Caffey complex, 0 to 3 percent slopes
69
Ruse mucky loam, rocky
72
Spot peat
84
Dawson-Kinross complex
90
Chippeny muck
91
Cathro and Lupton soils
93A
Ruse-Ensign complex, 0 to 3 percent slopes
120A
Charlevoix-Ensley complex, 0 to 3 percent slopes
122A
Wormet fine sandy loam, 0 to 3 percent slopes
249A
Iosco sand, 0 to 3 percent slopes
267A
Finch sand, 0 to 3 percent slopes
401A
Ingalls sand, 0 to 3 percent slopes
517B
Mancelona sand, 0 to 6 percent slopes
517D
Mancelona sand, 6 to 15 percent slopes
532
Angelica muck
541D
Kinross-Au Gres-Rubicon complex, 0 to 15 percent slopes

Habitat
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5

543B

Auger very fine sandy loam, sandy substratum, 0 to 6 percent slopes

5

553
564A
43
63
65
123
493A
516A
540D
222
W
66
498B

Carbondale-Loxley complex
Ingalls loamy sand, 0 to 3 percent slopes
Ensley muck
Deford muck
Ausable, Deford, and Tawas mucks, frequently flooded
Minocqua muck
Deford-Meehan-Seney complex, 0 to 3 percent slopes
Deford-Meehan complex, 0 to 3 percent slopes, drained
Deford-Rubicon-Au Gres complex, 0 to 15 percent slopes
Histosols and Aquents, ponded
Water
Markey mucky peat
Bursaw-Beavertail complex, 0 to 6 percent slopes

5
5
6
6
6
6
6
6
6
7
7
5
5

295

Appendix O:

Regression Analysis Output

296

Regression Analysis Output
_
Session Start: Friday, March 29th, 2013, 11:31:32 AM.
▼Binary Logistic Regression
The categorical values encountered during processing are
Variables
Levels
Random/Site (2 levels) Random
Habitat (8 levels)
Unclassified

Site
Jack Pine

Lowland Conifer

Lowland
Hardwood

Mixed Pine
Wetland/Marsh

Table 68: Categorical Variables

Categorical variables are dummy coded with the highest value as reference.
Dependent Variable
Input Records
Records for Analysis
Records Deleted for Missing Data
Table 69: Calculation Parameters

Dependent Variable Levels
Random
Site
Table 70: Sample Split

:
:
:
:

Random/Site
87
86
1

Category
Choices
RESPONSE
REFERENCE

Count
50
36

Failure to improve the likelihood function at Iteration 37
Old Log-Likelihood = -13.911 New Log-Likelihood = -13.911

297

Mixed Upland Northern
Hardwood

Log-Likelihood at Iteration1
-59.611
Log-Likelihood at Iteration2
-30.856
Log-Likelihood at Iteration3
-23.480
Log-Likelihood at Iteration4
-18.532
Log-Likelihood at Iteration5
-15.228
Log-Likelihood at Iteration6
-14.086
Log-Likelihood at Iteration7
-13.919
Log-Likelihood at Iteration8
-13.912
Log-Likelihood at Iteration9
-13.911
Log-Likelihood at Iteration10
-13.911
Log-Likelihood at Iteration11
-13.911
Log-Likelihood at Iteration12
-13.911
Log-Likelihood at Iteration13
-13.911
Log-Likelihood at Iteration14
-13.911
Log-Likelihood at Iteration15
-13.911
Log-Likelihood at Iteration16
-13.911
Log-Likelihood at Iteration17
-13.911
Log-Likelihood at Iteration18
-13.911
Log-Likelihood at Iteration19
-13.911
Log-Likelihood at Iteration20
-13.911
Log-Likelihood at Iteration21
-13.911
Log-Likelihood at Iteration22
-13.911
Log-Likelihood at Iteration23
-13.911
Log-Likelihood at Iteration24
-13.911
Log-Likelihood at Iteration25
-13.911
Log-Likelihood at Iteration26
-13.911
Log-Likelihood at Iteration27
-13.911
Log-Likelihood at Iteration28
-13.911
Log-Likelihood at Iteration29
-13.911
Log-Likelihood at Iteration30
-13.911
Log-Likelihood at Iteration31
-13.911
Log-Likelihood at Iteration32
-13.911
Log-Likelihood at Iteration33
-13.911
Log-Likelihood at Iteration34
-13.911
Log-Likelihood at Iteration35
-13.911
Log-Likelihood at Iteration36
-13.911
Log-Likelihood at Iteration37
-13.911
Log-Likelihood
-13.911
Table 71: Log-Likelihood Iteration History

AIC
45.821
Schwarz's BIC 67.911
Table 72: Information Criteria

298

Parameter

Estimate

CONSTANT
-2.480
Distance to Water
0.032
Habitat_Unclassified
19.798
Habitat_Jack Pine
-46.324
Habitat_Mixed Pine
-3.169
Habitat_Mixed Upland
-15.715
Habitat_Northern
1.343
Hardwood
Habitat_Lowland Conifer -0.183
Habitat_Lowland
37.708
Hardwood
Table 73: Parameter Estimates

Parameter

Standard Error Z

p-Value

1.375
0.009
1.198E+008
1.198E+008
1.770
110.119
1.607

-1.803
3.343
0.000
0.000
-1.791
-0.143
0.836

0.071
0.001
1.000
1.000
0.073
0.887
0.403

95% Confidence Interval
Lower
Upper
-5.176
0.216
0.013
0.050
-2.348E+008
2.348E+008
-2.348E+008
2.348E+008
-6.637
0.299
-231.544
200.113
-1.805
4.492

1.491
1.198E+008

-0.123
0.000

0.902
1.000

-3.106
-2.348E+008

Odds Ratio

Distance to Water
1.032
Habitat_Unclassified
3.964E+008
Habitat_Jack Pine
0.000
Habitat_Mixed Pine
0.042
Habitat_Mixed Upland
0.000
Habitat_Northern Hardwood
3.832
Habitat_Lowland Conifer
0.833
Habitat_Lowland Hardwood
2.379E+016
Table 74: Odds Ratio Estimates

Log-Likelihood of Constant Only Model
Log-Likelihood of Full Model
Chi-Square
df
p-Value
Table 75: Overall Model Fit

Standard Error
0.010
4.749E+016
0.000
0.074
0.000
6.156
1.242
2.851E+024

-58.466
-13.911
89.111
8
0.000

McFadden's Rho-Squared
0.762
Cox and Snell R-Square
0.645
Naglekerke's R-Square
0.868
Table 76: R-Square Measures

299

95% Confidence Interval
Lower
Upper
1.013
1.051
0.000
.
0.000
.
0.001
1.349
0.000
.
0.164
89.312
0.045
15.477
0.000
.

2.739
2.348E+008

Actual Choice

Predicted Choice
Actual Total
Response
Reference
Response
45.643
4.357
50.000
Reference
4.357
31.643
36.000
Predicted Total 50.000
36.000
86.000
Correct
0.913
0.879
Success Index
0.331
0.460
Total Correct
0.899
Table 77: Model Prediction Success Table

Sensitivity
Specificity
False Reference
False Response

0.913
0.879
0.087
0.121

Table 78: Summary of Prediction Success Table

Area under ROC
Curve

0.981

Figure 71: Receiver Operating Characteristic Curve

300

▼Logistic Regression: Simulation

Fixed Parameter
CONSTANT
Distance to Water
Habitat_Unclassified
Habitat_Jack Pine
Habitat_Mixed Pine
Habitat_Mixed Upland
Habitat_Northern Hardwood
Habitat_Lowland Conifer
Habitat_Lowland Hardwood
Table 79: Simulation Vector

Value
1.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000

Odds Ratio, 95.000 % Bounds = 0.084 [0.006, 1.241]

301

Appendix P:

Site Predictive Scores

302

Predictive
Rank
Score
20AR173/174
92
Medium
20AR245
155
High
20AR310
155
High
20AR338
140
High
20AR348
140
High
20AR350
140
High
20AR353
140
High
20AR358/386
108
Medium
20AR359
140
High
20AR398
140
High
20AR400
140
High
20AR406
9
Low
20AR435
140
High
20AR437
155
High
20AR495
108
Medium
20AR6
108
Medium
20CH2
108
Medium
20CH32
140
High
20CH433
140
High
20CH492
105
Medium
20CH86
98
Medium
20CH95
92
Medium
20DE106
155
High
20DE108
93
Medium
20DE167
108
Medium
20DE188
108
Medium
20DE236
59
Medium
20DE294
105
Medium
20DE296
93
Medium
20DE326
155
High
20DE378
108
Medium
20DE43
140
High
20DE459
140
High
20DE50
140
High
20DE75
140
High
20DE85
140
High
20DE93
105
Medium
20MK159
105
Medium
20MK24
59
Medium
20MK261
108
Medium
20MK3/11
108
Medium
Table 80: Predictive Scores HNF Sites
State No.

303

Table 80: (cont'd)
State No.
20MK334
20MK90
20ST109/110
20ST14
20ST227
20ST233
20ST262
20AR013

Predictive
Score
120
155
155
140
140
140
155
98

Rank
Medium
High
High
High
High
High
High
Medium

304

State No.

Predictive Score Rank

20AR210
98
Medium
20AR330
98
Medium
20CH171/172
123
Medium
20CH238
105
Medium
20CH27
98
Medium
20CH41
108
Medium
20CH43
98
Medium
20CH45
24
Low
20CH46
98
Medium
20CH6
42
Low
20CH77
94
Medium
20DE1
24
Low
20DE17
108
Medium
20DE19
94
Medium
20DE333
140
High
20DE4
105
Medium
20DE7
140
High
20MK1
108
Medium
20MK102
105
Medium
20MK169
105
Medium
20MK19
94
Medium
20MK22
140
High
20MK239
105
Medium
20MK375
105
Medium
20MK51/82/99
94
Medium
20MK53
105
Medium
20MK54
94
Medium
20MK58
9
Low
20MK6/7
108
Medium
20MK61
105
Medium
20ST1
105
Medium
20ST2
140
High
Table 81: Predictive Scores Other Sites

305

Appendix Q:

Site Data (LWPM and DUI Ranks)

309

State No.
Setting
LWPM Rank
20AR173/174
Interior
Medium
20AR210
Interior
Medium
20AR245
Interior
High
20AR310
Interior
High
20AR330
Coastal
Medium
20AR348
Coastal
High
20AR350
Coastal
High
20AR353
Coastal
High
20AR358/386
Coastal
Medium
20AR359
Coastal
High
20AR398
Coastal
High
20AR400
Coastal
High
20AR406
Coastal
Low
20AR435
Coastal
High
20AR437
Interior
High
20AR495
Coastal
Medium
20AR6
Coastal
Medium
20CH171/172
Interior
Medium
20CH2
Coastal
Medium
20CH238
Interior
Medium
20CH27
Coastal
Medium
20CH32
Coastal
High
20CH41
Interior
Medium
20CH43
Coastal
Medium
20CH433
Coastal
High
20CH45
Coastal
Low
20CH46
Coastal
Medium
20CH492
Coastal
Medium
20CH6
Coastal
Low
20CH77
Coastal
Medium
20CH86
Coastal
Medium
20CH95
Coastal
Medium
20DE106
Coastal
High
20DE108
Coastal
Medium
20DE167
Coastal
Medium
20DE17
Coastal
Medium
20DE188
Interior
Medium
20DE236
Coastal
Medium
20DE294
Interior
Medium
20DE296
Coastal
Medium
20DE326
Interior
High
20DE333
Coastal
High
20DE378
Interior
Medium
20DE4 (O/LW)
Coastal
Medium
20DE4 (pHST)
Coastal
Medium
20DE43
Interior
High
20DE459
Interior
High
20DE50
Interior
High
Table 82: Site Data (LWPM and DUI Ranks)

DUI Rank
Limited
Limited
Limited
Limited
Limited
Extended
Limited
Intermediate
Limited
Limited
Limited
Limited
Limited
Limited
Intermediate
Intermediate
Intermediate
Intermediate
Intermediate
Limited
Limited
Limited
Limited
Limited
Limited
Limited
Limited
Limited
Extended
Limited
Limited
Extended
Limited
Limited
Limited
Limited
Intermediate
Limited
Limited
Intermediate
Limited
Intermediate
Intermediate
Intermediate
Extended
Limited
Limited
Limited

307

Great Lake
Michigan
Michigan
Michigan
Michigan
Superior
Superior
Superior
Superior
Superior
Superior
Superior
Superior
Superior
Superior
Michigan
Superior
Superior
Huron
Superior
Huron
Superior
Superior
Superior
Huron
Superior
Huron
Huron
Superior
Huron
Superior
Superior
Superior
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan

Lake/River Early LW Late LW
River
Lake
River
x
Lake
x
x

x

x

x
x
x

Lake

x

Lake
x

x
x

x

x

River

River
x

x

x
x
x
x
x
x

x
x

x
x

River
x
River
Lake
River

Lake
Lake
River

x

x
x
x
x
x
x
x

Table 82: (cont'd)
State No.
Setting
20DE7
Coastal
20DE75
Interior
20DE85
Coastal
20DE93
Interior
20MK1 (Bois Blanc)Coastal
20MK1 (Juntunen) Coastal
20MK1 (Mackinac) Coastal
20MK102
Coastal
20MK159
Coastal
20MK169
Coastal
20MK19
Coastal
20MK22
Coastal
20MK239
Coastal
20MK24
Coastal
20MK261
Coastal
20MK3/11
Coastal
20MK334
Interior
20MK53
Coastal
20MK54
Coastal
20MK58
Coastal
20MK61
Coastal
20MK90
Coastal
20ST1
Coastal
20ST109/110
Interior
20ST14
Interior
20ST227
Interior
20ST233
Interior
20ST262
Interior

LWPM Rank
High
High
High
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Medium
High
Medium
Medium
Medium
Medium
Medium
Medium
Medium
Low
Medium
High
Medium
High
High
High
High
High

DUI Rank
Intermediate
Limited
Limited
Intermediate
Extended
Extended
Extended
Limited
Limited
Intermediate
Limited
Extended
Limited
Intermediate
Extended
Limited
Limited
Limited
Intermediate
Limited
Limited
Intermediate
Intermediate
Intermediate
Limited
Limited
Limited
Limited

308

Great Lake
Michigan
Michigan
Michigan
Michigan
Huron
Huron
Huron
Huron
Huron
Huron
Huron
Michigan
Huron
Michigan
Huron
Huron
Michigan
Huron
Huron
Huron
Huron
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan

Lake/River Early LW Late LW
River

x
x

River
x
x
x
x
x
x
x
x
x
x

x
x
x
x
x

Lake
x
x
x
x
Lake
Lake
River
Lake
Lake

x
x
x
x
x
x
x

Appendix R:

150 m Radius Catchment Data (SS&HD and DUI Ranks)

309

State No.
Setting
Great Lake
Lake/River
SS&HD Rank
20AR173/174 Interior
Interior
River
Moderate
20AR245
Interior
Interior
River
Increased
20AR310
Interior
Interior
Lake
Increased
20AR338
Coastal
Superior
Moderate
20AR348
Coastal
Superior
Moderate
20AR358/386 Coastal
Superior
Mimimal
20AR406
Coastal
Superior
Mimimal
20AR435
Coastal
Superior
Moderate
20AR437
Interior
Interior
Lake
Increased
20AR495
Coastal
Superior
Mimimal
20AR6
Coastal
Superior
Mimimal
20CH171/172 Interior
Interior
Lake
Increased
20CH2
Coastal
Superior
Mimimal
20CH32/433 Coastal
Superior
Increased
20CH492
Coastal
Superior
Mimimal
20CH86
Coastal
Superior
Mimimal
20CH95
Coastal
Superior
Mimimal
20DE106
Coastal
Michigan
Increased
20DE108
Coastal
Michigan
Mimimal
20DE167
Coastal
Michigan
Moderate
20DE188
Interior
Interior
River
Mimimal
20DE236
Coastal
Michigan
Moderate
20DE294
Interior
Interior
River
Moderate
20DE296
Coastal
Michigan
Moderate
20DE326
Interior
Interior
Lake
Increased
20DE378
Interior
Interior
River
Mimimal
20DE43
Interior
Interior
Lake
Moderate
20DE459
Interior
Interior
Lake
Moderate
20DE50
Interior
Interior
River
Increased
20DE75
Interior
Interior
River
Moderate
20DE85
Coastal
Michigan
Moderate
20DE93
Interior
Interior
River
Moderate
20MK159
Coastal
Huron
Moderate
20MK24
Coastal
Michigan
Mimimal
20MK261
Coastal
Huron
Mimimal
20MK3/11
Coastal
Huron
Mimimal
20MK334
Interior
Interior
Lake
Increased
20MK58
Coastal
Huron
Mimimal
20MK90
Coastal
Michigan
Increased
20ST109/110 Interior
Interior
Lake
Increased
20ST14
Interior
Interior
Lake
Moderate
20ST227
Interior
Interior
River
Moderate
20ST233
Interior
Interior
Lake
Moderate
20ST262
Interior
Interior
Lake
Increased
Table 83: 150 m Radius Catchment Data (SS&HD and DUI Ranks)

310

DUI Rank
Limited
Limited
Limited
Limited
Extended
Limited
Limited
Limited
Intermediate
Intermediate
Intermediate
Intermediate
Intermediate
Limited
Limited
Limited
Extended
Limited
Limited
Limited
Intermediate
Limited
Limited
Intermediate
Limited
Intermediate
Limited
Limited
Limited
Limited
Limited
Intermediate
Limited
Intermediate
Extended
Limited
Limited
Limited
Intermediate
Intermediate
Limited
Limited
Limited
Limited

Early LW

Late LW

x
x

x
x
x
x
x

x

x
x

x
x
x
x
x
x
x

x

x
x
x
x

x
x
x
x
x

x

x
x
x
x

Appendix S:

1500 m Radius Catchment Data (SS&HD and DUI Ranks)

311

State No.
Setting
Setting 2
SS&HD Rank
DUI Rank
20AR173/174
Interior
River
Moderate
Limited
20AR245&AR437
Interior
Lake
Increased
Intermediate
20AR310
Interior
Lake
Increased
Limited
20AR338 Cluster of 4 Coastal
Superior
Minimal
Limited
20AR348 Cluster of 4 Coastal
Superior
Minimal
Extended
20AR358&AR359
Coastal
Superior
Minimal
Limited
20AR435
Coastal
Superior
Moderate
Limited
20AR495
Coastal
Superior
Minimal
Intermediate
20CH171/172
Interior
Lake
Increased
Intermediate
20CH2
Coastal
Superior
Minimal
Intermediate
20CH32/433
Coastal
Superior
Moderate
Limited
20CH492
Coastal
Superior
Minimal
Limited
20CH86
Coastal
Superior
Minimal
Limited
20CH95
Coastal
Superior
Moderate
Extended
20DE106
Coastal
Michigan
Increased
Limited
20DE108
Coastal
Michigan
Minimal
Limited
20DE167 Cluster of 3 Interior
River
Increased
Intermediate
20DE188
Interior
River
Moderate
Intermediate
20DE236
Coastal
Michigan
Moderate
Limited
20DE296
Coastal
Michigan
Minimal
Intermediate
20DE326
Interior
Lake
Minimal
Limited
20DE378
Interior
River
Minimal
Intermediate
20DE43
Interior
Lake
Moderate
Limited
20DE459
Interior
Lake
Moderate
Limited
20DE50
Interior
River
Moderate
Limited
20DE75
Interior
River
Moderate
Limited
20DE85
Coastal
Michigan
Minimal
Limited
20MK159/261
Coastal
Huron
Minimal
Extended
20MK24
Coastal
Michigan
Increased
Intermediate
20MK3/11
Coastal
Huron
Moderate
Limited
20MK334
Interior
Lake
Minimal
Limited
20MK58/375
Coastal
Huron
Increased
Limited
20MK90
Coastal
Michigan
Increased
Intermediate
20ST109/110&ST262 Interior
Lake
Increased
Intermediate
20ST14&ST233
Interior
Lake
Increased
Limited
20ST227
Interior
River
Increased
Limited
Table 84: 1500 m Radius Catchment Data (SS&HD and DUI Ranks)

312

Early LW Late LW

x
x

x
x

x
x
x
x
x

x
x
x
x
x
x

x

x
x
x
x
x

x
x
x
x

x

x
x
x
x

Appendix T:

Fall and Spring Spawning Fish at Extended and Intermediate Diversity Sites

313

Site (Phase)
Basin/Region Fall Spawning Spring Spawning
20MK1 (Mackinac)
Straits
x
x
20MK22 (Mackinac)
Michigan
x
x
20MK261¹
Straits
x
20AR348
Superior
x
x
20CH6
Huron
x
20CH95²
Superior
x
20DE4 (proto-HST)
Michigan
20MK1 (Bois Blanc)
Straits
x
x
20MK1 (Juntunen)
Straits
x
x
20MK22 (Bois Blanc)
Michigan
x
x
20MK22 (Juntunen)
Michigan
x
x
¹=Whitefish family (Coregoninae) present, but not identified to species
²= Gill net sinkers recovered
Table 85: Extended Diversity Sites

Early/Late Sensitivity Big Game
M
E
x
H
E
x
M
E
H
E/L
x
L
E/L
x
M
L
x
L
M
x
M
L
x
M
L
x
H
L
x
H
L
x

Site (Phase)
Basin/Region Fall Spawning Spring Spawning Early/Late Sensitivity Big Game
20MK169/457 (Early)
Straits
M
x
x
E
x
20MK90
Straits
M
x
E
20DE4 (Oneota)
Bay de Noc
M
x
L
x
20DE296
Bay de Noc
M
x
x
L
x
20MK169/457 (Bois Blanc)
Straits
M
x
L
x
20AR359¹
Superior
H
x
x
L
20AR437
Interior
H
x
L
x
20DE75²
Interior
H
x
L
20DE188
Interior
M
x
L
20ST1
Michigan
M
x
x
E/L
x
20MK169/457 (Mixed)
Straits
M
x
x
E/L
x
20MK54
Straits
x
x
E/L
M
20MK61
M
x
x
E/L
¹=20AR359 is coded as limited diversity, but is part of a 150 m radius cluster with 20AR358 which is intermediate diversity
²=20DE75 is coded as limited diversity
Table 86: Intermediate Diversity Sites

314

Appendix U:

Sites Associated with Wild Rice Locales

315

State No.
LWSM Rank DUI Rank
Setting
20AR245
High
Limited
River
20AR310
High
Limited
Lake
20AR437
High
Intermediate Lake
20CH171/172 Medium
Intermediate Lake
20CH6
Low
Extended
Great Lake
20DE106
High
Limited
Great Lake
20DE326
High
Limited
Lake
20DE50
High
Limited
River
20MK334
Medium
Limited
Lake
20MK90
High
Intermediate Great Lake
20ST109/110 High
Intermediate Lake
20ST262
High
Limited
Lake
Table 87: Sites Associated with Wild Rice Locales

316

Early LW
x
x
-

Late LW
x
x
x
x
x
x
-

150 m Rank
Increased
Increased
Increased
Increased
Increased
Increased
Increased
Increased
Increased
Increased
Increased

1500 m Rank
Increased
Increased
Increased
Increased
Increased
Minimal
Moderate
Minimal
Increased
Increased
Increased

Appendix V:

Sites in Mixed Pine Habitats

317

State No.
Setting
LWPM Rank
20AR245
Interior
High
20AR310
Interior
High
20AR348
Coastal
High
20AR350
Coastal
High
20AR353
Coastal
High
20AR359
Coastal
High
20AR398
Coastal
High
20AR400
Coastal
High
20AR435
Coastal
High
20AR437
Interior
High
20CH32
Coastal
High
20CH433
Coastal
High
20DE106
Coastal
High
20DE236
Coastal
Medium
20DE326
Interior
High
20DE333
Coastal
High
20DE43
Interior
High
20DE459
Interior
High
20DE50
Interior
High
20DE7
Coastal
High
20DE75
Interior
High
20DE85
Coastal
High
20MK22
Coastal
High
20MK24
Coastal
Medium
20MK90
Coastal
High
20ST109/110 Interior
High
20ST14
Interior
High
20ST227
Interior
High
20ST233
Interior
High
20ST262
Interior
High
Table 88: Sites in Mixed Pine Habitats

DUI Rank
Limited
Limited
Extended
Limited
Intermediate
Limited
Limited
Limited
Limited
Intermediate
Limited
Limited
Limited
Limited
Limited
Intermediate
Limited
Limited
Limited
Intermediate
Limited
Limited
Extended
Intermediate
Intermediate
Intermediate
Limited
Limited
Limited
Limited

318

Lake
Michigan
Michigan
Superior
Superior
Superior
Superior
Superior
Superior
Superior
Michigan
Superior
Superior
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan

Early LW Late LW
x
x
x
x

x

x

x
x

x
x

x
x
x
x

x
x
x

x
x
x
x
x
x

Appendix W:

Extended Diversity Sites

319

State No.
Great Lake
20AR348
Superior
20CH6
Huron
20CH95
Superior
20DE4 (Protohistoric) Michigan
20MK1 (Bois Blanc)
Huron
20MK1 (Juntunen)
Huron
20MK1 (Mackinac)
Huron
20MK22
Michigan
20MK261
Huron
Table 89: Extended Diversity Sites

LWPM Rank
High
Low
Medium
Medium
Medium
Medium
Medium
High
Medium

DUI Rank Early LW Late LW
Extended
x
x
Extended
x
x
Extended
x
Extended
x
Extended
x
Extended
x
Extended
x
Extended
x
x
Extended
x

320

Occupations
Late Archaic - Contact
Middle Woodland - Contact
Middle Woodland - Late Woodland
Middle Woodland - Contact
Middle Woodland - Contact
Middle Woodland - Contact
Middle Woodland - Contact
Multiple Late Woodland Components
Middle Woodland - Early Late Woodland

Appendix X:

Ceramics Data

321

State Number Early Type
20AR245
20AR348
Madison
20AR358/386
20AR359
20AR437
20CH2
Mackinac
20CH32
Mackinac
20CH43
20CH45
20CH492
20CH6
Mackinac
20CH95
20DE108
20DE17
Mackinac
20DE236
20DE296
20DE326
20DE333
20DE4 (O/LW)
20DE7
Heins Creek
20DE75
20MK1
Mackinac
20MK102
Mackinac
20MK159
Mackinac
20MK169/457 Mackinac
20MK19
Mackinac
20MK22
Mackinac
20MK239
20MK24
Spring Creek
20MK261
Mackinac
20MK53
20MK54
Mackinac
20MK61
Mackinac
20MK90
Mackinac
20ST1
20ST109/110
20ST14
Table 90: Ceramics Data

Late Type
Point Sauble
Sand Point
Sand Point
Sand Point
Oneota
Bois Blanc
Bois Blanc
Juntunen
Juntunen
Iroquoian
Sand Point
Sand Point
Oneota
Sand Point
Oneota
Oneota
Oneota
Oneota
Juntunen
Juntunen
Iroquoian
Oneota
Juntunen

Juntunen
Juntunen
Juntunen
Oneota
Oneota
Oneota

Setting
Interior
Coastal
Coastal
Coastal
Interior
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Interior
Coastal
Coastal
Coastal
Interior
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Coastal
Interior
Interior

LWPM Rank
High
High
Medium
High
High
Medium
High
Medium
Low
Medium
Low
Medium
Medium
Medium
Medium
Medium
High
High
Medium
High
High
Medium
Medium
Medium
Medium
Medium
High
Medium
Medium
Medium
Medium
Medium
Medium
High
Medium
High
High

322

FDUI Rank
Limited
Extended
Limited
Limited
Intermediate
Intermediate
Limited
Limited
Limited
Limited
Extended
Extended
Limited
Limited
Limited
Intermediate
Limited
Intermediate
Extended
Intermediate
Limited
Extended
Limited
Limited
Intermediate
Limited
Extended
Limited
Intermediate
Extended
Limited
Intermediate
Limited
Intermediate
Intermediate
Intermediate
Limited

Great Lake
Michigan
Superior
Superior
Superior
Michigan
Superior
Superior
Huron
Huron
Superior
Huron
Superior
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Michigan
Huron
Huron
Huron
Huron
Huron
Michigan
Huron
Michigan
Huron
Huron
Huron
Huron
Michigan
Michigan
Michigan
Michigan

Locality
Indian River
Grand Island
Grand Island
Grand Island
Indian River
Whitefish Bay
Whitefish Bay
N. Lake Huron
N. Lake Huron
Whitefish Bay
N. Lake Huron
Whitefish Bay
Bay de Noc
Bay de Noc
Bay de Noc
Bay de Noc
Indian River
Bay de Noc
Bay de Noc
Bay de Noc
Sturgeon River
Mackinac Straits
Mackinac Straits
Mackinac Straits
Mackinac Straits
Mackinac Straits
N. Lake Michigan
Mackinac Straits
Mackinac Straits
Mackinac Straits
Mackinac Straits
Mackinac Straits
Mackinac Straits
Mackinac Straits
N. Lake Michigan
Indian River
Indian River

BIBLIOGRAPHY

323

BIBLIOGRAPHY

Abrams, M.
1992 Fire and the Development of Oak Forests. BioScience 42(5):346-353.
Abrams, M. D. and G. J. Nowacki
2008 Native Americans as active and passive promoters of mast and fruit trees in the eastern
USA. The Holocene 18(7):1123-1137.
Aiken, S. G., P. F. Lee, D. Punter, and J. M. Stewart
1988 Wild Rice in Canada. NC Press Ltd, Toronto.
Albert, D. A.
1995 Regional Landscape Ecosystems of Michigan Minnesota, and Wisconsin: A Working
Map and Classification. Gen. Tech. Rep. NC-178. USDA Forest Service, North Central
Forest Experiment Station, St. Paul.
2003 Between Land and Lake: Michigan’s Great Lakes Coastal Wetlands. Michigan Natural
Features Inventory. Lansing, Michigan.
Albert, D. A. and P. Comer
2008 Atlas of Early Michigan’s Forests, Grasslands, and Wetlands: An Interpretation of the
1816‐1856 General Land Office Surveys. Michigan State University Press, East Lansing.
Albert, D. and L. Minc
1987 The Natural Ecology and Cultural History of the Colonial Point Red Oak Stands.
University of Michigan Biological Research Station, Michigan Department of Natural
Resources, Lansing, Michigan.
Alverson, W., D. Waller, S. Solheim
1988 Forests Too Deer: Edge Effects in northern Wisconsin. Conservation Biology 2(4):348358.
Anderton, J. B.
1993 Paleoshoreline Geoarchaeology in the Northern Great Lakes, Hiawatha National Forest.
Heritage Monograph No. 1. USDA-Forest Service, Hiawatha National Forest, Escanaba,
Michigan
1995 Paleoshoreline Geoarchaeology in the Northern Great Lakes: An Interdisciplinary
Approach to Prehistoric Coastal Settlement. Unpublished Doctoral Dissertation,
Department of Geography, University of Wisconsin, Madison, Wisconsin
1999 Native American, Fire-Maintained Blueberry Patches in the Coastal Pine Forests of the
Northern Great Lakes. The Great Lakes Geographer 6 (1&2):29-39
324

Anderton, J. B., S. B. Dunham, and M. C. Branstner
1995 1994 Cultural Resource Surveys: Hiawatha National Forest. Great Lakes Research
Associates, Inc., Williamston, Michigan.
Anderton, J. B., R. Legg, and R. Regis
2009 Geoarchaeological Approaches to Site Location Modeling and Archaeological Survey in
the Pictured Rocks National Lakeshore, Michigan USA. The Wisconsin Archeologist
90:19-30
2011 Geoarchaeological Inventory of Specific Lakeshore Landforms at Pictured Rocks
National Lakeshore. Department of Geography, Northern Michigan University,
Marquette, Michigan
Anderton, J. B., D. J. Weir, and G. G. Robinson
1991 1990 Cultural Resource Survey: Hiawatha National Forest. Commonwealth Cultural
Resources Group, Inc., Jackson, Michigan
Andrews, W. L.
2011 Archaeological Data Recovery of ‘Getewaaking’: A Middle to Late Woodland and
Historic Period Site (20MK451) Located on Mackinac Island, Mackinac County,
Michigan. Andrews Cultural Resources, Ann Arbor, Michigan.
Anschuetz, K., R. Wilshusen, and C. Scheick
2001 An Archaeology of Landscapes: Perspectives and Directions. Journal of Archaeological
Research 9(2):157-211.
Arbogast , A. F.
2009 Sand Dunes. In Michigan Geography and Geology, edited by R. Schaetzl, J. Dudent. And
D. Brandt, pp. 274-288. Custom Publishing, New York.
Arbogast, A. F., A. G. Wintle, and S. C. Packman
2002 Widespread middle-Holocene dune formation in the eastern Upper Peninsula of Michigan
and the relationship to climate and outlet-controlled lake level. Geology 30(1):55-58.
Arzigian, C.
2000 Middle Woodland and Oneota Contexts for Wild rice Exploitation in Southwestern
Wisconsin. Midcontinental Journal of Archaeology 25(2):245-268.
Baedke, S. J. and T. A. Thompson
2000 A 4,700-Year Record of Lake Level and Isostasy for Lake Michigan. Journal of Great
Lakes Research 26(4):416-426.
Baerreis, D. A. and J. E. Freeman
1958 Late Woodland Pottery in Wisconsin as Seen from Aztalan. The Wisconsin Archeologist
39:36-61.

325

Bailey, G.
2007 Time Perspectives, Palimpsests, and the Archaeology of Time. Journal of
Anthropological Archaeology 26(3):198-233.
Baker, R. H.
1983 Michigan Mammals. Michigan State University Press, East Laning.
Becker, G. C.
1983 Fishes of Wisconsin. University of Wisconsin Press, Madison.
Bernabo, J.C.
1981 Quantitative estimates of temperature changes over the last 2700 years in Michigan based
on pollen data. Quaternary Research 15, 143–59.
Bettinger, R. L.
1987 Archaeological Approaches to Hunter-Gatherers. Annual Review of Anthropology
16:121-142.
Bianchi, T. H.
1974 Description and Analysis of the Prehistoric Materials Recovered at the Winter Site.
Unpublished Master’s Thesis, Department of Anthropology, Western Michigan
University, Kalamazoo.
Bigony, B.
1968 Archaeological Survey of the Southern Shore of Lake Superior. On file at the University
of Michigan Museum of Anthropology, Ann Arbor, Michigan.
Binford, L. R.
1980 Willow Smoke and Dogs’ Tails: Hunter-Gatherer Settlement Systems and Archaeological
Site Formation. American Antiquity. 45(1):4-20.
1983 In Pursuit of the Past: Decoding the Archaeological Record. Thames and Hudson, New
York.
2001 Constructing Frames of Reference: An Analytical Method for Archaeological Theory
Building Using Hunter-Gatherers and Environmental Data Sets. University of California
Press, Berkeley.
Black, B., and M. Abrams
2001 Influences of Native Americans and Surveyor Biases on Metes and Bounds Witness-Tree
Distribution. Ecology 82(9):2547-2586.
Blackbird, A. J.
1897 Complete Both Early and Late History of the Ottawa and Chippewa Indians of Michigan.
Babcock and Darling, Harbor Springs, Michigan.

326

Booth, R., S. Jackson, and C. Gray
2004 Paleoecology and high-resolution paleohydrology of a kettle peatland in upper Michigan.
Quaternary Research 61 (2004) 1 – 13
Bourdo, E. A.
1954 A Validation of Methods Used in Analyzing Original Forest Cover. Unpublished
Doctoral Dissertation, University of Michigan, Ann Arbor.
Brandt, K. L.
1996 The Effects of Early Agriculture on Native North American Populations: Evidence from
the Teeth and Skeleton. Unpublished Ph.D. dissertation, Department of Anthropology,
University of Michigan, Ann Arbor.
Branstner, C. N.
1995 Archaeological Investigations at the Cloudman Site (20CH6): A Multicomonent Native
American Occupation on Drummond Island, Michigan. Department of Anthropology,
Michigan State University, East Lansing, Michigan.
Branstner, M. C., M. Hambacher, M. Holman
2000 1998 Cultural Resource Surveys and Test Excavations: Hiawatha National Forest. Part
II – Phase II Evaluations. Great Lakes Research Associates, Inc., Williamston, Michigan.
Branstner, S. M.
1991 Decision-Making Processes in a Culture Contact Context : The Case of the Tionontate
Huron of the Upper Great Lakes. Unpublished Ph.D. dissertation, Department of
Anthropology, Michigan State University, East Lansing.
Brashler, J. G., E. B. Garland, M. B. Holman, W. A. Lovis, and S. R. Martin
2000 Adaptive Strategies and Socioeconomic Systems in Northern Great Lakes Riverine
Environments: The Late Woodland of Michigan. In Late Woodland Societies: Tradition
and Transformation across the Midcontinent, edited by T. E. Emerson, D. L. McElrath,
A. C. Fortier, pp. 543-579. University of Nebraska Press, Lincoln, Nebraska.
Braun, D. P. and S. Plog
1982 Evolution of “Tribal” Social Networks: Theory and Prehistoric North American
Evidence. American Antiquity 47(3):504-525.
Brewer, R., G. McPeek, and R. Adams
1991 The Atlas of Breeding Birds of Michigan. Michigan State University Press, East Lansing,
Michigan.
Brose, D. S.
1970 The Archaeology of Summer Island: Changing Settlement Systems in Northern Lake
Michigan. Anthropological Papers 41. Museum of Anthropology, University of
Michigan, Ann Arbor.

327

Brose, D. S. and M. J. Hambacher
1999 The Middle Woodland in Northern Michigan. In Retrieving Michigan’s Buried Past: The
Archaeology of the Great Lakes State, edited by J. R. Halsey, pp. 173-192. Cranbrook
Institute of Science, Bulletin 64. Bloomfield Hills, Michigan.
Brown, J. A.
1985 Long-Term Trends to Sedentism and the Emergence of Complexity in the American
Midwest. In Prehistoric Hunter-Gatherers: The Emergence of Cultural Complexity,
edited by T. D. Price and J. A. Brown, pp. 201-231. Academic Press.
Brown, T.
1999 1999 Wild Rice Lake and Stream Survey. Bay Mills Indian Community, Brimley,
Michigan
Brubaker, L.
1975 Postglacial Forest Patterns Associated with Tills and Outwash in Northcentral Upper
Michigan. Quaternary Research 5:499-527.
Brugam, R., M. Giordi, C. Sesvold, S. Johnson, R. Almos
1997 Holocene Vegetation History in the Sylvania Wilderness Area of the Western Upper
Peninsula of Michigan. American Midlands Naturalist 137(1):62-71.
Bruhy, M. E. and K. Egan-Bruhy
2009 Preliminary Report of the Investigation of Archibald Lake Mounds and Village
(47 Oc-309): A Wolf River Tradition Settlement. Paper Presented at the Midwest
Archaeological Conference, Iowa City, Iowa.
Buckmaster, M. M.
1979 Woodland and Oneota Settlement and Subsistence Systems in the Menominee River
Watershed. Unpublished Ph.D. dissertation, Department of Anthropology, Michigan
State University, East Lansing
1983 A Phase II Archaeological Evaluation of the Sturgeon River Site (20DE75), Delta
County, Michigan. Northern Michigan University Archaeological Laboratory, Marquette,
Michigan.
2004 The Northern Limits of Ridge Field Agriculture: An Example from Menominee County.
In An Upper Great Lakes Archaeological Odyssey: Essays in Honor of Charles E.
Cleland, edited by W. A. Lovis, pp. 30-42. Cranbrook Institute of Science, Bloomfield
Hills, Michigan.
Buol, S. W., F. Hole and R. McCracken
1989 Soil Genesis and Classification. Iowa State University Press, Ames.

328

Burger, T. L. and J. Kotar
2003 A Guide to Forest Communities and Habitat Types of Michigan. The Department of
Forest Ecology and Management, University of Wisconsin-Madison, Madison.
Burrough, P. A. and R. A. McDonnell
1998 Principles of Geographic Information Systems. Oxford University Press, New York.
Bursey, J. A.
2001 Storage Behavior in the Northeast: A Review of the Evidence. North American
Archaeologist. 22(3)179-199.
Clark, C. P.
1993 Archeological Survey and Sitte Testing at Pictured Rocks National Lakeshore, Alger
County, Michigan. Midwest Archaeological Center Technical Report No. 23, U.S
National Park Service, Lincoln, Nebraska.
Cleland, C. E.
1966 The Prehistoric Animal Ecology and Ethnozoology of the Upper Great Lakes Region.
Anthropological Papers No. 29. Museum of Anthropology, University of Michigan, Ann
Arbor.
1982 The Inland Shore Fishery of the Northern Great Lakes: Its Development and Importance
in Prehistory. American Antiquity 47:761-784.
1983 Indians in a Changing Environment. In The Great Lakes Forest, An Environmental and
Social History, edited by S. Flader, pp. 83-95. University of Minnesota Press,
Minneapolis.
1989 Comments on “A Reconsideration of Aboriginal Fishing Strategies in the Northern Great
Lakes Region” by Susan R. Martin. American Antiquity 54(3):605-609.
1992a Rites of Conquest: The History and Culture of Michigan’s Native Americans. University
of Michigan Press, Ann Arbor.
1992b From Ethnohistory to Archaeology: Ottawa and Ojibwa Band Territories of the Northern
Great Lakes. In Text Aided Archaeology, edited by B. Little, pp. 97-102. Telford Press,
Caldwell, New Jersey.
Cleland, C. E. (editor)
1971 The Lasanen Site: An Historical Burial Locality in Mackinac County, Michigan.
Publications of the Museum, Anthropological Series 1. Michigan State University, East
Lansing.

329

Cleland, C. E. and G. R. Peske
1968 The Spider Cave Site. In The Prehistory of the Burnt Bluff Area, edited by J. Fitting,
pp. 20-60. Anthropological Papers 34. University of Michigan Museum of Anthropology,
Ann Arbor.
Coffman, M., E. Alyanak, J. Kotar, J. Ferris
1980 Field Guide Habitat Classification System for the Upper Peninsula of Michigan and
Northeast Wisconsin. CROFS; Department of Forestry, Michigan Technological
University, Houghton, Michigan.
Comer, P., D. Albert, H. Wells, B. Hart, J. Raab, D. Price, D. Kashian, and R. Corner
1995 Michigan’s Presettlement Vegetation, as Interpreted from the General Land Office
Surveys 1816-1856. Michigan Natural Features Inventory, Lansing, Michigan.
Cooper, J.
1996 Cloudman Site (20CH6), Drummond Island, Michigan, Features 26 and 27, 1992
Excavations. On file at the Consortium for Archaeological Research, Department of
Anthropology, Michigan State University, East Lansing.
Cordain, L., J. Miller, S. Eaton, N. Mann, S. Holt, and J. Speth
2000 Plant-Animal Subsistence Ratios and Macronutrient Energy Estimations in Worldwide
Hunter-Gatherer Diets. American Journal of Clinical Nutrition 71:682-692.
Cremin, W. M.
1980
Sand Point (20BG14): A Lakes Phase Site on the Keweenaw Peninsula, Baraga
County, Michigan. Michigan Archaeologist 26:7-16.
Crow, T.
1988 Reproductive Mode and Mechanisms for Self-Replacement of northern Red Oak
(Quercus rubra)-A Review. Forest Science 34(1)19-40.
Cunningham, P.
2011 Caching your Savings: The Use of Small-Scale Storage in European Prehistory. Journal
of Anthropological Archaeology 30:135-144.
Dalla Bona, L.
2000 Protecting Cultural Resources through Forest Management Planning in Ontario Using
Archaeological Predictive Modeling. In Practical Applications of GIS for
Archaeologists: A Predictive Modeling Toolkit, edited by K. Wescott and R. J. Brandon,
pgs 73-99. Taylor and Francis, Inc., Philadelphia
Davidson-Hunt, I.
2003 Indigenous Land Management, Cultural Landscapes, and Anishnaabe People of Shoal
Lake, Northwestern Ontario, Canada. Environments 31(1):21-41.

330

Davidson-Hunt, I. and F. Berkes
2003 Learning as You Journey: Anishnaabe Perception of Social-Ecological Environments
and Adaptive Learning. Conservation Ecology 8(1):5 Online resource (last accessed
1/3/2011): http://www.consecol.org/vol8/iss1/art5
Davis, M., C. Douglas, R. Calcote, K. Cole, M. Winkler, R. Flakne
2000 Holocene Climate in the Western Great Lakes National Parks and Lakeshores:
Implications for Future Climate Change. Conservation Biology 14(4):968-983.
Delcourt, H., and P. Delcourt
1996 Presettlement Landscape Heterogeneity: Evaluating Grains of Resolution Using General
Land Office Survey Data. Landscape Ecology 11(6):363-381.
Delcourt, P. and H. Delcourt
2004 Prehistoric Native Americans and Ecological Change: Human Ecosystems in Eastern
North America since the Pleistocene. Cambridge University Press, Cambridge.
Delcourt, P., P. Nester, H. Delcourt, C. Mora, and K. Orvis
2002 Holocene Lake-Effect Precipitation in Northern Michigan. Quaternary Research 57:225223.
Demeritt, D.
1991 Agriculture, Climate, and Cultural Adaptation in the Prehistoric Northeast. Archaeology
of Eastern North America 19:183-202
Demers, P. A.
1991 Report on the St. Marys River Archaeological Survey: Archaeological Reconnaissance
and Limited Testing on the Lower St. Marys River, Chippewa County, Michigan. On file
at the Department of Anthropology, Michigan State University, East Lansing, Michigan.
Densmore, F.
1974 How Indians Use Wild Plants for Food, Medicine and Crafts (formerly Uses of Plants by
Chippewa Indians). Dover Publications, Incorporated, New York.
Dewar, R. A.
1986 Discovering Settlement Systems of the Past in New England Site Distributions. Man in
the Northeast 31:77-88.
Dey, D.
2002 Fire History and Postsettlement Distrubance. In Oak Forest Ecosystems: Ecology and
Management for Wildlife, W. McShea and W. Healy (eds.), pgs 46-59, The John Hopkins
University Press, Baltimore, Maryland.
Dodge, R. L.
1973 Michigan Ghost Towns, Volume IIII, the Upper Peninsula. Glendon Publishing
Company, Sterling Heights, Michigan.
331

Dorney, J.
1981 The Impact of Native Americans on Presettlement Vegetation in Southeastern Wisconsin.
Wisconsin Academy of Sciences, Arts, and Letters 69:26-36.
Dorney, C. and J. Dorney
1989 An Unusual Oak Savannah in Northeastern Wisconsin: The Effects of Indian-Caused
Fire. American Midlands Naturalist 122(1):103-113.
Dorothy, L. G.
1980 The Ceramics of the Sand Point Site (20BG14), Baraga County, Michigan: A Preliminary
Description. The Michigan Archaeologist 26:39-90
Dorr, J. A. and D. E. Eschman
1986 Geology of Michigan. University of Michigan Press, Ann Arbor.
Dorwin, J., R. Guendling, and A. Brine
1980 A Cultural Resource Survey of Proposed Undertakings in the Hiawatha National Forest:
Alger, Delta, Schoolcraft, Chippewa, and Mackinac Counties, Michigan. Soil Systems,
Inc., Bloomington, Indiana.
Drake, E. and S. Dunham
2004 The Woodland Period Occupation of Grand Island. Midcontinental Journal of
Archaeology 29(2):133-166.
Drake, E., J. Franzen, and J. Skibo
2009 Chronological Patterns of Lithic Raw Material Choice in the Grand Island-Munising Bay
Locality of Michigan’s Upper Peninsula: Archaeological Implications for Inter- and
Intra-site Comparisons. The Wisconsin Archeologist 90(1&2):132-149
Drake, M. and S. Dunham
2007 2005 Cultural Resource Surveys: Hiawatha National Forest. Commonwealth Cultural
Resources Group, Inc., Jackson, Michigan.
Dunham, S. B.
2000a Cache Pits: Archaeology, Ethnohistory and Continuity of Tradition. In The Contribution
of Archaeology and Material Culture Studies to Ethnohistory, edited by Michael
Nassaney and Eric Johnson, pp. 225-260. Society for Historical Archaeology and
University Press of Florida, Gainesville.
2000b 1999 Cultural Resource Surveys and Test Excavations: Hiawatha National Forest.
Commonwealth Cultural Resources Group, Inc., Jackson, Michigan.

332

2002 Appendix D: Preliminary Comments on the Settlement and Subsistence Patterns for the
Initial and Terminal Woodland Periods in the Eastern Upper Peninsula of Michigan. In
S. B. Dunham and M. J. Hambacher, 2000 Cultural Resource Evaluations: Hiawatha
National Forest. Commonwealth Cultural Resources Group, Inc. Jackson, Michigan.
2008 Wild Rice in the Eastern Upper Peninsula: A Review of the Evidence. Paper presented at
the 2008 Annual Meetings of the Michigan Academy of Science, Arts & Letters, Botany
and Plant Ecology Division, Kalamazoo, Michigan.
2009 Nuts about Acorns: A Pilot Study on Acorn Use in Woodland Period Subsistence in the
Eastern Upper Peninsula of Michigan. The Wisconsin Archeologist 90(1-2):113-130.
2012 Late Woodland Landscapes in the Eastern Upper Peninsula of Michigan. Paper
presented at the Annual Meetings of the Society for American Archaeology, Memphis,
Tennessee.
2013 2012 Cultural Resource Surveys: Hiawatha National Forest, Michigan. Commonwealth
Cultural Resources Group, Inc., Jackson, Michigan
Dunham, S. B. and J. B. Anderton
1999 Late Archaic Radiocarbon Dates from the Popper Site (FS 09-10-03-825/20AR350): A
Multicomponent Site on Grand Island, Michigan. The Michigan Archaeologist 45:1-22.
Dunham, S. B. and M. Branstner
1993 1992 Cultural Resource Surveys: Hiawatha National Forest. Great Lakes Research
Associates, Inc., Williamston,
1995 1994 Phase II Cultural Resource Evaluations: Hiawatha National Forest. Great Lakes
Research Associates, Inc., Williamston, Michigan.
1998 1997 Cultural Resource Surveys and Test Excavations: Hiawatha National Forest. Great
Lakes Research Associates, Williamston, Michigan.
Dunham, S. B., M. Drake, M. Jeakle, and M. Hambacher
2010 2009 Cultural Resource Surveys and Test Excavations: Hiawatha National Forest.
Commonwealth Cultural Resources Group, Inc., Jackson, Michigan.
Dunham, S. B. and M. Hambacher
2002 Cultural Resource Evaluations: Hiawatha National Forest. Commonwealth Cultural
Resources Group, Inc., Jackson, Michigan.
2007 Cultural Resource Evaluation: Bark Dock Site (FS 09-10-04-012/20CH95), Hiawatha
National Forest, Chippewa County, Michigan. Commonwealth Cultural Resources
Group, Inc., Jackson, Michigan.

333

Dunham, S. B. and M. Jeakle
2012 2011 Cultural Resource Surveys: Hiawatha National Forest, Michigan. Commonwealth
Cultural Resources Group, Inc., Jackson, Michigan.
Dunham, S. B., M. Hambacher and M. Branstner
1993 1992 Phase II Cultural Resource Evaluations: Hiawatha National Forest. Great Lakes
Research Associates, Inc., Williamston, Michigan.
1994 1993 Cultural Resource Surveys an Test Excavations: Hiawatha National Forest. Great
Lakes Research Associates, Inc., Williamston, Michigan.
1997 1996 Cultural Resource Surveys and Test Excavations: Hiawatha National Forest. Great
Lakes Research Associates, Williamston, Michigan.
Dunham, S. B., M. Drake, M. Jeakle, and M. Hambacher
2010 2009 Cultural Resource Surveys and Test Excavations: Hiawatha National Forest.
Commonwealth Cultural Resources Group, Inc., Jackson, Michigan.
Dunham, S., K. Kaufmann, K. Hagenmaier, S. Norby, M. Jeakle, and K. Taylor
2011 Archaeological Investigations of the Back Forty Joint Venture, Back Forty Project Area,
Menominee County, Michigan. Commonwealth Cultural Resources Group, Inc., Jackson,
Michigan.
Ebert, D.
2004 Predictive Modeling and the Ecology of Hunter-Gatherers of the Boreal Forest of
Manitoba. British Archaeological Reports, BAR International Series 1221,
Archaeopress, Oxford, England:
Edman, F. R.
1969 Wild Rice in Michigan. In A Study of Wild Rice in Minnesota, F. R. Edman (ed.), pgs. 3338, Minnesota Resources Commission, St. Paul.
Egan-Bruhy, K. C.
2007 20CH6. Cloudman Site Floral Table. On file at the Consortium for Archaeological
Research, Department of Anthropology, Michigan State University, East Lansing.
Eichenlaub, V., J. Harmon, F. Nurnberger, H. Stolle
1990 The Climatic Atlas of Michigan. University of Notre Dame Press, South Bend, Indiana.
Fitting, J. E.
1968a Northern Lake Michigan Lithic Industries. In Prehistory of the Burnt Bluff Area, edited
by J. E. Fitting, pp. 116-133. Anthropological Papers 34. University of Michigan
Museum of Anthropology, Ann Arbor.

334

1968b The Spring Creek Site. In Contributions to Michigan Archaeology, edited by J. E. Fitting,
J. R. Halsey, and H. M. Wobst, pp. 1-178. Anthropological Papers No. 32, Museum f
Anthropolog, Univ erity of Michigan, Ann Arbor.
1975a The Archaeology of Michigan: A Guide to the Prehistory of the Great Lakes Region.
Bulletin 56. Cranbrook Institute of Science, Bloomfield Hills, Michigan.
1975b Appendix I: Aboriginal Artifacts. In Excavations at the Schoolcraft House, prepared by A
Hugley and D Basler. Manuscript on file to at the Office of the State Archaeologist,
Lansing, Michigan.
1978 The Gyftakis Site and McGregor Sites: Middle Woodland Occupations in St. Ignace,
Michigan. Manuscript on file at the Office of the State Archaeologist, Lansing,
Michigan.
1980 Te Oskonchiae. Manuscript on file at the Office of the State Archaeologist, Lansing,
Michigan.
Fitting, J. E. (ed)
1974 Contributions to the Archaeology of the St. Ignace Area. The Michigan Archaeologist 20
(3-4):117-286.
1976 Archaeological Excavations at the Marquette Mission Site, St. Ignace, Michigan, in 1972.
The Michigan Archaeologist 22(2-3):103-282.
Fitting, J. E. and W. S. Clarke
1974 The Beyer Site (SIS 20). The Michigan Archaeologist 20:227-277.
Fitting, J.E., and C. E. Cleland
1969 Late Prehistoric Settlement Patterns in the Upper Great Lakes. Ethnohistory 16:289-302.
Fitting, J. E. and K. A. Cushman
1974 The Sposito Site (SIS-29). The Michigan Archaeologist 20(3-4):171-194.
Fitting, J. E. and P. L. Fisher
1975 And Archaeological Survey of the St. Ignace Area. In Archaeological Site Survey in the
Mackinac Straits, edited by L. M. Stone, pp. 5-90. Mackinac Island State Park
Commission, Mackinac City, Michigan.
Fitting, J. E. and M. J. Lynott
1974 The Norge Vilage Site (SIS 16). The Michigan Archaeologist 20 (3-4):195-226.
Flannery, K. V.
1968 Archaeological Systems Theory and Early Mesoamerica. In Anthropological
Archaeology in the Americas, edited by B. J. Meggars, pp. 67-87. Anthropological
Society of Washington, Washington, D.C.
335

Franzen, J. G.
1975 An Archaeological Survey of Chippewa County, Michigan. Michigan History Division
Archaeological Survey Report, Lansing, Michigan.
1979 A Cultural Resource Survey of Portions f the Hiawatha National Forest, Alger,
Chippewa, Delta, Mackinac, and Schoolcraft Counties, Michigan, Commonealth
Associates, Inc., Jackson, Michigan.
1983 1982 Cultural Resource Survey and Test Excavation, Hiawatha National Forest.
Cultural Resource Management Report No. 2. USDA Forest Service, Hiawatha National
Forest, Escanaba, Michigan.
1986 Prehistoric Settlement on the Hiawatha National Forest, Michigan: A Preliminary
Locational Model. Cultural Resource Management Report 4. USDA Forest Service,
Hiawatha National Forest, Escanaba, Michigan.
1987 Test Excavation and Locational Analysis of Prehistoric Sites in the Hiawatha National
Forest, Michigan: 1985 Season. Cultural Resource Management Report 5. USDA
Forest Service, Hiawatha National Forest, Escanaba, Michigan.
1998 Archaeological Test Excavation, Monitoring, and Surface Collection, Hiawatha National
Forest, Michigan. Heritage Resource Program Report No. 10. Hiawatha National
Forest, Escanaba, Michigan.
2000 Preliminary Report on Archaeological Investigations at the Williams Trading Post Site,
FS 09-10-03-811 (20AR353), Grand Island National Recreation Area, Alger County,
Michigan. Hiawatha National Forest, Escanaba, Michigan.
Franzen, J. G. and E. C. Drake
2005 Test Excavations at Kitchi Minissing (Grand Island): Evaluation of Five Archaeological
Sites on the Shore of Lake Superior. Hiawatha National Forest, Escanaba, Michigan.
Funk, C. L.
2004 Optimal Foraging Theory and Cognitive Archaeology: Cup’ik Cultural Perception in
Southwestern Alaska. In Hunters and Gatherers in Theory and Archaeology, edited by
G. M. Crothers, pp. 279-298. Center for Archaeological Investigations, Southern Illinois
University, Carbondale
Futyma, R. P.
1982 Postglacial Vegetation of Eastern Upper Michigan (Volumes I and II). Unpublished
Ph.D. Dissertation, Department of Botany, University of Michigan, Ann Arbor.

336

Gallagher, J. P. and C. M. Arzigian
1994 A New Perspective on Late Prehistoric Agricultural Intensification in the Upper
Mississippi River Valley. In Agricultural Origins and Development in the Midcontinent,
edited by W. Green, pp. 171-184. University of Iowa Office of the State Archaeologist,
Iowa City.
Gallagher, J. P., R. F. Boszhardt, R. F. Sasso, and K. Stevenson
1985 Oneota Ridged Field Agriculture in Southwestern Wisconsin. American Antiquity
50:605-612.
Gardner, P. S.
1997 The Ecological Structure and Behavioral Implications of Mast Exploitation Strategies. In
Peoples, Plants, and Landscapes: Studies in Paleoethnobotany, edited by K. J.
Gremillion, pp 161-178, University of Alabama Press, Tuscaloosa.
Gibbon, G. E.
2002 Archaeological Predictive Modeling: An Overview. In A Predictive Model of
Precontact Archaeological Site Location for the State of Minnesota, edited by G. J.
Hudak. E. Hobbs, A. Brooks, C. Seersland, and C. Phillips, Appendix A. Minneosta
Department of Transportation, St. Paul, MN. Online resource (last accessed 1/3/2011):
http://www.mnmodel.dot.state.mn.us/chapters/app_a.htm
Gibbon, G. and C. Caine
1980 The Middle and Late Woodland Transition in Eastern Minnesota. Midcontinenal Jounal
Archaeology 5(1):57-72.
Goltz, G. E.
1992 1991 Cultural Resource Survey, Manistique, Munising, and Rapid River Ranger District:
Hiawatha National Forest. Leech Lake Heritage Sites Program, Cass Lake, Minnesota.
Grove, M.
2009 Hunter-gatherer movement patterns: Causes and constraints. Journal of Anthropological
Archaeology 28:222-233.
Grunwald, Sabine
2013 Mullisols.
http://soils.ifas.ufl.edu/faculty/grunwald/teaching/eSoilScience/mollisols.shtml#Mollisols
_distinguish (accessed April 24, 2013)
Hambacher, M. J., and M. B, Holman.
1995 Camp, Cache and Carry: The Porter Creek South Site (20MN100) and Cache Pits at
20MN31 in the Manistee National Forest. The Michigan Archaeologist 41(2-3):47-94.
Hart, J. and W. Lovis
2013 Reevaluating What We Know About the Histories of Maize in Northeastern North
America: A Review of Current Evidence. Journal of Archaeological Research
21(2):175-216.
337

Harpending, H. and H. Davis
1977 Some Implications for Hunter-Gatherer Ecology Derived from the Spatial Structures of
Resources. World Archaeology 8(3):275-286.
Hiawatha National Forest (HNF)
1938 Lake and Stream Survey. On file at the Hiawatha National Forest, Gladstone, Michigan.
nd

Cultural Resource Inventory Form; Site FS 09-10-01-076. On file at the Hiawatha
National Forest, Gladstone, Michigan.

Hickerson, H.
1960 The Feast of the Dead among the Sevententh-Century Algonkians of the Upper Great
Lakes. American Anthropologist 62:81-107.
1962 The Southwestern Chippewa: An Ethnohistorical Study. Memoir 92, American
Anthropological Association, Vol. 64, No. 3, Part 2.
Hildebrand, E. A.
2003 Motives and opportunities for domestication: an ethnoarchaeological study in southwest
Ethiopia. Journal of Anthropological Archaeology 22:358-375.
Holling, C. S.
1973 Resilience and Stability of Ecological Systems. Annual Review of Ecology and
Systematics 4:1-23.
Holman, M. B.
1978 The Settlement System of the Mackinac Phase. Unpublished Ph.D. dissertation,
Department of Anthropology, Michigan State University, East Lansing.
Holman, M. B. and R. G. Kingsley
1996 Territoriality and Social Interaction During the Early Late Woodland Period in Southern
Michigan. In Investigating the Archaeological Record of the Great Lakes State: Essays
in Honor of Elizabeth Baldwin Garland, edited by M. B. Holman, J. G. Brashler, and K.
E. Parker, pp. 341-382. New Issues Press, Western Michigan University, Kalamazoo,
Michigan.
Holman, M. B. and F. Krist
2001 Late Woodland Storage and Mobility in Western Lower Michigan. The Wisconsin
Archeologist 82(1&2):7-32.
Holman, M. B. and W. A. Lovis
2008 The Social and Environmental Constraints on Mobility in the Late Prehistoric Upper
Great Lakes Region. In The Archaeology of Mobility: Old World and New World
Nomadism, edited by H. Bernard and W. Wandrich. Cotsen Institute of Archaeology,
UCLA, Los Angeles
338

Howey, M. C. L.
2006 Ritual, Resources, and Regional Organization in the Upper Great Lakes, AD 1200-1600.
Unpublished Ph.D. Dissertation, Department of Anthropology, University of Michigan,
Ann Arbor.
Howey, M. C. L. and K. Parker
2008 Camp, Cache, Stay Awhile: Preliminary Considerations of the Social and Economic
Process of Cache Pits along Douglas Lake, Michigan. The Michigan Archaeologist
54:19-44.
Ingold, T.
1983 The Significance of Storage in Hunting Societies. Man 18(3):553-571
Janzen, D. E.
1968 The Naomikong Point Site and the Development of Laurel in the Lake Superior Basin.
Anthropological Papers 36. University of Michigan Museum of Anthropology, Ann
Arbor.
Jenks, A. E.
1900 The Wild Rice Gatherers of the Upper Great Lakes: A Study in American Primitive
Economics. Nineteenth Annual Report of the Bureau of American Ethnology for the
Years 1897-1898, Part 2, Pages 1013-1137. Government Printing Office, Washington
DC.
Jenness, D.
1935 The Ojibwa Indians of Parry Island, Their Social and Religious Life. National Museum
of Canada, Bulletin No. 78, Anthropological Series No. 117, Ottawa
Jochim, M. A.
1976 Hunter-Gatherer Subsistence and Settlement: A Predictive Model. Academic Press, New
York.
1991 Archaeology as Long Term Ethnography. American Anthropologist 93(2):308-321.
Johnson, E.
1969 Preliminary Notes on the Prehistoric Use of Wild Rice. Minnesota Archaeologist
30(2):31-43.
Johnson, L. M.
2000 “A Place That’s Good,” Gitksan Landscape Perception and Ethnoecology. Human
Ecology 28(2):301-325.

339

Jones, B. A.
1993 Archaeological Inventory and Evaluative Testing in Pictured Rocks National Lakeshore,
Michigan, 1985-1990. Midwest Archaeological Center, Occasional Studies I
Anthropology No. 30, National Park Service, Lincoln, Nebraska.
Jones, E. E.
2010 An Analysis of Factors Influencing Sixteenth and Seventeenth Century Haudenosaunee
(Iroquois) Settlement Locations. Journal of Anthropological Archaeology 29:1-14.
Justice, N. D.
1987 Stone Age Spear and Arrow Points of the of the Midcontinental and Eastern United
States. Indiana University Press, Bloomington.
Kapp, R. O.
1999 Michigan Late Pleistocene, Holocene, and Presettlement Vegetation and Climate. In
Retrieving Michigan’s Buried Past: The Archaeology of the Great Lakes State, edited by
J. R. Halsey, pp. 31-58. Cranbrook Institute of Science, Bulletin 64. Bloomfield Hills,
Michigan.
Katzenberg, M. A.
1989 Stable Isotope Analysis of Archaeological Faunal Remains from Southern Ontario.
Journal of Archaeological Science 16:319-329.
Keeley, L. H.
1995 Protoagricultural practices among hunter-gatherers: A cross-cultural survey. In Last
Hunters – First Farmers: New Perspectives on the Prehistoric Transition to Agriculture,
edited by T. D. Price and A. B. Gebauer, pp. 243-72. School of American Research Press,
Santa Fe, NM.
Kelly, R. L.
1992 Mobility/Sedentism: Concepts, Archaeological Measures, and Effects. Annual Review of
Anthropology. 21: 43-66.
1995 The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Smithsonian
Institution Press, Washington.
Kinietz, V. W.
1965 Indians of the Western Great Lakes, 1615-1760. Ann Arbor Paperbacks, Ann Arbor
Kintigh, K. W.
1984 Measuring Archaeological Diversity by Comparison with Simulated Assemblages.
American Antiquity 49(1):44-54.
Kohler, T. A. and S. C. Parker
1986 Predictive Models for Archaeological Resource Location. In Advances in Archaeological
Method and Theory, edited by M. B. Schiffer, pp. 397-452. Academic Press, New York.
340

Kooiman, S. M.
2012 Old Pots, New Approaches: A Functional Analysis of Woodland Pottery from Lake
Superior’s South Shore. Unpublished Master’s Thesis, Department of Anthropology,
Illinois State University, Normal.
Kooyman, B. P.
2000 Understanding Stone Tools and Archaeological Sites. University of New Mexico Press,
Albuquerque.
Kost, M., D. Albert, J. Cohen, B. Slaughter, R. Schillo, C. Weber, and K. Chapman.
2007 Natural Communities of Michigan: Classification and Description. Michigan Natural
Features Inventory, Report Number 2007-21, Lansing, Michigan.
Kotar, J.
1976 Soil-Habitat Type Relationships in Michigan and Wisconsin. Journal of Soil and Water
Conservation 41(5):347-349.
Kuhnlien, Harriet V. and Nancy J. Turner.
1991 Traditional Plant Foods of Canadian Indigenous Peoples: Nutrition, Botany, and Use.
Gordon and Breach Science Publishers, Philadelphia, Pennsylvania
Kvamme, K. L.
1985 Determining Empirical Relationships between the Natural Environment and Prehistoric
Site Locations: A Hunter-Gatherer Example. In For Concordance in Archaeological
Analysis:Bridging Data Structure, Quantitative Technique, and Theory, edited by C.
Carr, pp. 208-238. Westport Press, Kansas City.
1992 A Predictive Site Location Model on the High Plains: An Example with an Independent
Test. Plains Anthropologist 37(138):19-40.
Larson, C. E.
1999 A Century of Great Lakes Levels Research. In Retrieving Michigan’s Buried Past: The
Archaeology of the Great Lakes State, edited by J. R. Halsey, pp. 1-30. Cranbrook
Institute of Science, Bulletin 64. Bloomfield Hills, Michigan.
Leahy, M. and K. Pregitzer
2003 A Comparison of Presettlement and Present-Day Forests in Northeastern Lower
Michigan. American Midland Naturalist 149(1):71-89.
Lofstrom, T.
1987 The Rise of Wild Rice Exploitation and its Implication fr Population Size and Social
Organization in Minnesota Woodland Period Cultures. The Minnesota Archaeologist
46(2):3-15.
Loope, W. and J. Anderton
1998 Human vs. Lightning Ignition of Presettlement Surface Fires in Coastal Pine Forests of
the Upper Great Lakes. American Midland Naturalist 140(2):206-218.
341

Loope, W. L., T. Fisher, H. Jol, R. Goble, J. Anderton, W. Blewett
2004 A Holocene history of dune-mediated landscape change along the southeastern shore of
Lake Superior. Geomorphology 61:303-322.
Lovis, W. A.
1976 Quarter Sections and Forests: An Example of Probability Sampling in the Northeastern
Woodlands. American Antiquity 41:364-372.
1979 A Cultural Resource Management Study of the Hiawatha National Forest, Michigan:
Phase II, the Hypothesis Tests. Archaeological Survey Report 36. Michigan State
University Museum, East Lansing.
2008 Revisiting the Johnson Site (20CN46): A Winter Camp in a Cedar Swamp on Mullet
Lake, Michigan. The Michigan Archaeologist 54:73-92.
Lovis, W. A. and M. Holman
1976 Subsistence Strategies and Population: A Hypothetical Model for the Development of
Late Woodland in the Straits of Mackinac/Sault Ste. Marie Area. The Michigan
Academician 8(3):267-276.
Lovis, W. A., A. F. Arbogast, and G. W. Monaghan
2012 The Geoarchaeology of Lake Michigan Coastal Dunes. Environmental Research Series,
No. 2, J. A. Robertson series editor. Michigan Department of Transportation and
Michigan State University Press, East Lansing.
Lovis, W., R. Donahue, and M. Holman
2005 Long-Distance Logistic Mobility as an Organizing Principle Among Northern HunterGatherers: A Great Lakes Middle Holocene Settlement System. American Antiquity
70(4):660-693.
Lovis, W. A., T. W. Langhorne, and S. Van Arsdel
1978 A Cultural Resource Overview of the Huron-Manistee National Forest. Michigan State
University Museum, East Lansing.
Lovis, W. A., G. W. Monaghan, K. C. Egan, and B. A. Smith
2001 Wetlands and Emergent Horticultural Economies in the Upper Great Lakes: A New
Perspective from the Schultz Site. American Antiquity 66:615-632.
Lovis, W. A. G. W. Monaghan, A. F. Arbogast, and S. L. Foreman
2012 Differential Spatial Preservation of Archaeological Sites in a Great Lakes Coastal Zone.
American Antiquity 77(3):591-608.
Lovis, W. A., G. W. Monaghan, K. C. Egan, and B. A. Smith
2001 Wetlands and Emergent Horticultural Economies in the Upper Great Lakes: A New
Perspective from the Schultz Site. American Antiquity 66:615-632.

342

Lu, Y. and D. Waller.
1996 Genetic Variability in Wild Rice (Zinzania palustris va. palustris) populations in
Northern Wisconsin. Great Lakes Indian Fish and Wildlife Commision, Odanah,
Wisconsin. Department of Botany, University of Wisconsin-Madison
Luedke, B. E.
1976 Lithic Material Distributions and Interaction Patterns during the Lat Woodland Period
in Michigan. Unpublished Ph.D. Dissertation, Department of Anthroplogy, University of
Michigan, Ann Arbor.
Lynott, M. J.
1974 The Ferrier Site. The Michigan Archaeologist 20(3-4):139-170.
Malainey, M. E.
2007 Fatty Acid Analysis of Archaeological Residues: Procedures and Possibilities. In Theory
and Practice of Archaeological Residue Analysis, edited by H. Barnard and J. W.
Eerkens, pp.77-89. British Archaeological Reports International Series, Oxford.
Martin, S. R.
1977 A Preliminary Cultural Resource Management Study of the Hiawatha National Forest,
Michigan. Archaeological Survey Report No. 20. Michigan State University Museum,
East Lansing.
1979 Final Report: Phase II Archaeological Site Evaluation of the Gros Cap Cemetery Area,
Mackinac County, Michigan. Cultural Resource Report No. 6, Michigan Technological
University, Houghton, Michigan.
1985 Models of Change in the Woodland Settlement of the Northern Great Lakes Region.
Unpublished Doctoral thesis, Department of Anthropology, Michigan State University,
East Lansing.
1989 A Reconsideration of Aboriginal Fishing Strategies in the Northern Great Lakes Region.
American Antiquity 54:594-604.
1999 A Site for All Seasons: Some Aspects of Life in the Upper Peninsula During Late
Woodland Times. In Retrieving Michigan’s Buried Past: The Archaeology of the Great
Lakes State, edited by J. R. Halsey, pp. 221-227. Bulletin 64, Cranbrook Institute of
Science, Bloomfield Hills, Michigan.
Martin, S. R. and P. E. Martin
1980 An Archaeological Survey of the Proposed Route of US-2 Between Thompson and Rapid
River, Michigan. Cultural Resource Management Report No. 9, Michigan Technological
University, Houghton, Michigan.

343

Martin, T. J.
1980 Animal Remains form 20MK90 (Pointe aux Chenes), Mackinac County, Michigan.
Manuscript on File at the Archaeological Laboratory, Michigan Technological
University, Houghton.
1982 Animal Remains from the Scott Point Site: Evidence for Changing Subsistence Strategies
during the Late Woodland Period in Northern Michigan. Paper Presented at the Annual
Meeting of the Michigan Academy of Science, Arts, and Letters, Kalamazoo, Michigan.
1991 Faunal Analysis. In 1990 Cultural Resource Survey: Hiawatha National Forest, edited by
J. Anderton, D. Weir, and G. Robinson, pp. 195-200. Commonwealth Cultural Resources
Group, Inc., Jackson, Michigan
Martin, T. J. and A. R. Perri
2011 Animal Remains from the Getewaaking Site (20MK457), a Muliple-Component Middle to
Late Woodland Habitation Site on Mackinac Island, Michigan. Technical Report 2011000-7, Illinois State Musum, Landscape History Program, Springfield.
Martin, T. J., J. B. Anderton, and J. G. Franzen
1993 Northern Exposure: Late Woodland Environment and Subsistence in the Hiawatha
National Forest of Northern Michigan. Paper presented at the Annual Meetings of the
Society for American Archaeology, St. Louis, Missouri.
Mayry, K. J.
1995 Interim Report on Archaeological Investigations on the Starr-Mayry Site: 1995 Field
Season. Manuscript on file at the Office of the state Archaeologist, Lansing, Michigan.
Mason, R. J.
1966 Two Stratified Sites on the Door Peninsula of Wisconsin. Anthropological Papers No. 26.
University of Michigan Museum of Anthropology, Ann Arbor.
1981 Great Lakes Archaeology. Academic Press, New York.
Mather, D. and R. G. Thompson
2000 Archaeological Perspectives on Wild Rice. In Proceedings of the Wild Rice Research and
Management Conference, edited by L. S. Williamson, L. A. Dlutkowski, and A. P. Soltis,
pgs 15-26, Great Lakes Indian Fish and Wildlife Commission, Odonah, Wisconsin.
McElrath, D., T. Emerson, and A. Fortier
2000 Social Evolution or Social Response? A Look at the “Good Gray Cultures” After Four
Decades of Midwest Research. In Late Woodland Societies: Tradition and
Transformation across the Midcontinent, edited by T. E. Emerson, D. L. McElrath, A. C.
Fortier, pp. 3-36. University of Nebraska Press, Lincoln, Nebraska.

344

McHale-Milner, C.
1991 Localization in Small-Scale Societies: Late Prehistoric Social Organization in the
Western Great Lakes. In Between Bands and States, edited by Susan A. Gregg, pp. 35-57.
Center for Archaeological Investigations, Occasional Paper No. 9. Southern Illinois
University
1998 Ceramic Style, Social Differentiation, and Resource uncertainty in the Late Prehistoric
Upper Great Lakes. Unpublished Doctoral Dissertation, Department of Anthropology,
University of Michigan, Ann Arbor.
McKern, W. C.
1931 Wisconsin Pottery. American Anthropologist 33(3): 383-389
McPherron, A.
1967 The Juntunen Site and the Late Woodland Prehistory of the Upper Great Lakes Area.
Anthropological Papers No. 30. University of Michigan Museum of Anthropology, Ann
Arbor.
Michigan Department of Natural Resources (MDNR)
c.1949 Survey of Wild Rice Planting in Michigan. Online Resource,
http://ww2.dnr.state.mi.us/publications/pdfs/huntingwildlifehabitat/reports/wldlibrary/1000-1099/1022.pdf (last accessed 11/12/13)
Michigan Geographic Data Library (MiGDL)
2013a Great Lakes Shoreline Geographic Theme
http://www.mcgi.state.mi.us/mgdl/?rel=thext&action=thmname&cid=3&cat=Great+Lake
s+Shoreline (last accessed 11/12/13)
2013b Lake Polygons Geographic Theme
http://www.mcgi.state.mi.us/mgdl/?rel=thext&action=thmname&cid=3&cat=Lake+Polyg
ons (last accessed 11/12/13)
2013c MI Geographic Framework Hydrography Lines (v13a) Geographic Theme
http://www.mcgi.state.mi.us/mgdl/?rel=thext&action=thmname&cid=3&cat=MI+Geogra
phic+Framework+Hydrography+Lines+%28v13a%29 (last accessed 11/12/13)
2013d Digital Elevation Model (DEM) Geographic Theme
http://www.mcgi.state.mi.us/mgdl/?rel=thext&action=thmname&cid=13&cat=Digital+El
evation+Model+%28DEM%29 (last accessed 11/12/13)
2013e Land Cover Circa 1800 Geographic Theme
http://www.mcgi.state.mi.us/mgdl/?rel=thext&action=thmname&cid=5&cat=Land+Cove
r+Circa+1800 (last accessed 11/12/13)

345

2013f SSURGO Soils Geographic Theme
http://www.mcgi.state.mi.us/mgdl/?rel=thext&action=thmname&cid=12&cat=SSURGO
+Soils (last accessed 11/12/13)
Michigan State University Herbarium (MSU Herbarium)
nd
Records on file, MSU Herbarium, East Lansing, Michigan
Miller, A. M. and I. Davidson-Hunt
2010 Fire, Agency, and Scale in the Creation of Aboriginal Cultural Landscapes. Human
Ecology 38:401-414.
Miller, H.
1943 Wild Rice in Michigan. Michigan Conservation June:4-5
Minc, L. D.
1986 Scarcity and Survival: The Role of Oral Tradition in Mediating Subsistence Crisis.
Journal of Anthropological Archaeology 5:39-113
Moffat, C. R.
1979 Some Observations on the Distribution and Significance of the Garden Beds in
Wisconsin. The Wisconsin Archeologist 60:222-248.
Moffat C. R. and C. M. Arzigian
2000 New Data on Late Woodland Use of Wild Rice in Northern Wisconsin.
Journal of Archaeology 25(1):49-82.

Midcontinental

Monckton, S. G.
1992 Huron Paleoethnobotany. Ontario Archaeological Reports 1. Ontario Heritage
Foundation, Toronto, Ontario
Morgan, C.
2009 Climate change, uncertainty and prehistoric hunter-gather mobility. Journal of
Anthropological Archaeology 28:382-396.
2012 Modeling Modes of Hunter-Gatherer Food Storage. American Antiquity 77(4):714-736.
Moore, C. R. and V. Thompson
2012 Animism and Green River Persistent Places: A Dwelling Perspective of the Shell Mound
Archaic. Journal of Social Archaeology 12(2):264-284.
Mulholland, S. C.
2000 Analysis of Phytoliths from Sediments, 20ME61, Michigan. Report No. 00-12.
Archaeometry Laboratory, University of Minnesota, Duluth.

346

Nelson, M., M. Hegmon, S. Kulow, and K. Schollmeyer
2006 Archaeological and Ecological Perspectives on Reorganization: A Case Study from the
Mimbres Region of the U.S. Southwest. American Antiquity 71(3):403-432.
Nicholas, G. P.
2006 Prehistoric Hunter-Gatherers in Wetland Environments: Theoretical Issues, Economics
Organization and Resource Management Strategies. In Wetlands: Local Issues and
World Perspectives, edited by M. Lillie and S. Ellis, pp. 46-62. Oxbow Press, Oxford,
U.K..
Nowacki, G. J. and Abrams, M. D.
2008. The demise of fire and “mesophication” of forests in Eastern United States. BioScience
58: 123-138.
Nowacki, G., M. Abrams, C. Lorimer
1990 Composition, Structure, and Historical Development of Northern Red Oak Stands along
an Edaphic Gradient in North-Central Wisconsin. Forest Science 36(2):276-292.
Office of the State Archaeologist, Michigan (OSA)
nda
20DE333 Archaeological Site File. On file at the Office of the State Archaeologist,
Lansing, Michigan.
ndb

20ST2 Archaeological Site File. On file at the Office of the State Archaeologist, Lansing,
Michigan.

O’Gorman, J.
2007 The Myth of Mocassin Bluff – Rethinking the Potawatomi Pattern. Ethnohistory
54(3):373:406.
O’Shea, J. M.
1981 Coping with Scarcity: Exchange and Social Storage. In Economic Archaeology:
Towards an Integration of Ecological and Social Approaches, edited by A. Sheridan and
G. Bailey. B.A.R. International Series 96, Oxford
1989 The role of wild resources in small scale agricultural systems: tales from the Lakes and
the Plains. In Bad Year Economics: Cultural Responses to Risk and Uncertainty, edited
by P. Halstead and J. M. O’Shea, pp. 57-67. Cambridge University Press, New York.
2003 Inland Foragers and the Adoption of Maize Agriculture in the Upper Great Lakes of
North America. Before Farming: The Archaeology and Anthropology of HunterGatherers 2003/1(3):68-83.
O’Shea, J. and P. Halstead
1989 Conclusions: Bad Year Economics. In Bad Year Economics: Cultural Responses to Risk
and Uncertainty, edited by P. Halstead and J. O’Shea, pp. 123-126. Cambridge
University Press, New York.
347

O’Shea, J. M. and C. McHale-Milner
2002 Material Indicators of Territory, Identity, an Interaction in a Prehistoric Tribal System. In
The Archaeology of Tribal Societies, Edited by W. A. Parkinson, pp. 200-226.
International Mongraphs in Prehistory, Ann Arbor, Michigan
Overstreet, D. F.
1997 Oneota History and Prehistory. Wisconsin Archeologist 78:250-297.
2009 The Mero Complex and the Menominee Tribe – Prospects for Territorial
Ethnicity. The Wisconsin Archeologist 90(1&2):132-177.
Parker, K. E.
1996 Three Corn Kernels and a Hill of Beans: The Evidence for Prehistoric Horticulture in
Michigan. In Investigating the Archaeological Record of the Great Lakes State: Essays
in Honor of Elizabeth Baldwin Garland, edited by M. B. Holman, J. G. Brashler, and K.
E. Parker, pp. 341-382. New Issues Press, Western Michigan University, Kalamazoo,
Michigan.
Parkinson, W. A.
2002 Introduction: Archaeology and Tribal Societies. In The Archaeology of Tribal Societies,
edited by W. A. Parkinson, pp. 1-12, International Monographs in Prehistory, Ann Arbor,
Michigan.
Peters, J. H.
1986 Transmission Line Planning and Archaeological Resources: A Model of Archaeological
Potential for Southwestern Ontario. In Archaeological Consulting in Ontario: Papers of
the London Conference, 1985, edited by W. A. Fox. Occasional Papers No. 2. Ontario
Archaeological Society, London Chapter.
Pirnie, M. D.
1935 Michigan Waterfowl Management. Michigan Department of Conservation, Lansing,
Michigan
Politis, G. G.
1996 Moving to Produce: Nukak Mobility and Settlement Patterns in Amazonia. World
Archaeology 27(3):492-511.
Prahl, E. J.
1986 Archaeological Testing in Marquette Park, Mackinac Island, Michigan: An Addendum to
the Mackinac Island Sewer and Watermain Report. Caminos Associates, Bay City,
Michigan.

348

Price, D. L.
1994 An ecological study of the composition, structure and disturbance regimes of the preEuropean settlement forests of western Chippewa County, Michigan. Unpublished
Masters Thesis, Department of Forestry, Michigan State University, East Lansing.
Quimby, George I.
1962 A Year with a Chippewa Family, 1764-1764. Ethnohistory 9(3):217-239.
Raymond, J. S., and W. DeBoer
2006 Maize on the Move. In Histories of Maize: Multidisciplinary Approaches to the
Prehistory, Linguistics, Biogeography, Domestication, and Evolution of Maize, edited by
J. Stahler, R. Tykot, B. Benz, pgs 337-342, Academic Press, New York.
Redman, C. L.
2005 Resilience Theory in Archaeology. American Anthropologist 107(1):70-77.
Redman, C. L. and A. P. Kinzig
2003 Resilience of Past Landscapes: Resilience Theory, Society, and the Longue Durée.
Conservation Ecology 7(1):14. Online resource (last accessed 1/3/2011):
http://www.consecol.org/vol7/iss1/art14
Richner, J. J.
1973 Depositional History and Tool Industries at the Winter Site: A Lake Forest Middle
Woodland Cultural Adaptation. Unpublished Masters thesis, Department of
Anthropology, Western Michigan University, Kalamazoo.
Rist, S. G.
2008 Legacies of Forest Management and Fire in Mixed-Pine Forest Ecosystems of the Seney
National Wildlife Refuge, Eastern Upper Michigan. Unpublished Master’s Thesis,
Environment and Natural Resource Program, the Ohio State University, Columbus.
Ritchie, J. C.
1986 Climate Change and Vegetation Response. Vegetatio 67:65-74.
Robinson, G., D. J. Weir, and M. J. Hambacher
1991 1990 Cultural Resource Survey of Parcels on Grand Island Hiawatha National Forest.
Commonwealth Cultural Resource Group, Inc., Jackson, Michigan.
Robertson, J. A.
1993 The Lithic Assemblage from Sites 20BY77, 20BY78, and 20BY79, Bay City, Michigan.
In The Archaic, Woodland and Historic Occupations of the Liberty Bridge Locale, Bay
City, Michigan, edited by W. A. Lovis, pp. 41-97. Michigan Cultural Resource
Investigation Series 3. Lansing, Michigan.

349

Robertson, J. W. Lovis, and J. Halsey
1999 The Late Archaic: Hunter Gatherers in an Uncertain Environment. In Retrieving
Michigan’s Buried Past: The Archaeology of the Great Lakes State, edited by John R.
Halsey, pp. 125-146. Cranbrook Institute of Science, Bulletin 64. Bloomfield Hills,
Michigan.
Rogers E. S, and M. B. Rogers
1976 Subsistence Strategies in the Fish and Hare Period, Northern Ontario: The Weagaw
Ojibwa, 1880-1920. Journal of Anthropological Research 32(1):1-42.
Roper, D. C.
1979 The Method and Theory of Site Catchment Analysis: A Review. Advances in
Archeological Method and Theory 2:119-140.
Ruffner, C. and M. Abrams
2002 Dendrochronological Investigation of Disturbance History for a Native American Site in
Northwestern Pennsylvania. Journal of the Torrey Botanical Society 129(3):251-260.

Rutter, W. and D. Weir
1985 1984 Cultural Resource Survey and Test Excavations: Hiawatha National Forest.
Gilbert/Commonwealth Inc., Jackson, Michigan.
1986 1985 Cultural Resources Survey and Text Excavations: Hiawatha National Forest.
Gilbert/Commonwealth, Inc., Jackson, Michigan.
1988 1987 Cultural Resource Survey and Test Excavations: Hiawatha National Forest.
Gilbert/Commonwealth Inc., Jackson, Michigan.
1989 1988 Cultural Resource Survey and Test Excavations: Hiawatha National Forest.
Gilbert/Commonwealth, Inc., Jackson, Michigan.
1990 1989 Cultural Resource Survey and Test Excavations: Hiawatha National Forest.
Commonwealth Cultural Resources Group, Inc., Jackson, Michigan.
Rutter, W. E., D. J. Weir, and D. C. Roper
1984 1983 Cultural Resource Survey of the Hiawatha National Forest. Commonwealth
Associates, Inc., Jackson, Michigan.
Ryden, N. L., L. Goldstein, and E. D. Benchley
1983 Final Report on the Initial Phases of Predictive Model Development for Aboriginal
Cultural resources on the Chequamegon National Forest, Wisconsin. Report of
Investigations 73. University of Wisconsin-Milwaukee Archaeological Research
Laboratory, Milwaukee.
Sasso, R. F.
2001 Observations on the Oneota Cultivation Sites in Winnebago County. The
Wisconsin Archeologist 82:125-138.
350

Scarry, C. M.
2003 Patterns of Wild Plant Utilization in the Prehistoric Eastern Woodlands. In People and
Plants in Eastern North America, edited by P. E. Minnis, pp. 50-104. Smithsonian
Books, Washington DC.
Schaetzl, R. J., H. Enander, M. Luehmann, D. Lusch, C. Fish, M. Bigsby, M. Stelgmeyer, J.
Gorsco, C. Fragees, and A. Pollyea
2013 Mapping the Physiography of Michigan with GIS. Physical Geography 34(1):2-39.
Schalk, R. F.
1977 The Structure of Anadromous Fish Resource. In For Theory Building in Archaeology,
edited by L. R. Binford, pp. 207-243. Academic Press, New York.
Schroeder, S.
2004 Current Research on Late Pre-Contact Societies of the Midcontinental United States.
Journal of Archaeological Research 12(4):311-372.
Schlanger, S. H.
1992 Recognizing Persistent Places in Anasazi Settlement Systems. In Space, Time, and
Archaeological Landscapes, edited by J. Rossignol and L. Wandsnider, pp. 91-112.
Plenum Press, New York.
Schoolcraft, H. R.
1966 Travels Through the Northwestern Regions of the United States. University Microfilms,
Ann Arbor, Michigan.
1975 Personal Memoirs of a Residence of Thirty Years with the Indian Tribes on the American
Frontier. Arno Press, New York.
Seeman, M. F. and W. S. Dancey
2000 The Late Woodland Period in Southern Ohio: Basic Issues and Prospects. In Late
Woodland Societies: Tradition and Transformation across the Midcontinent, edited by T.
E. Emerson, D. L. McElrath, A. C. Fortier, pp. 583-612. University of Nebraska Press,
Lincoln, Nebraska.
Seymour, D. J.
2010 Sanctioned Inequity and Accessibility Issues in the Grey Literature in the United States.
Archaeologies: Journal of the World Archaeological Congress 6(2):233-269.
Skibo, J. M., M. Malainey, and E. Drake
2009 Stone Boiling, Fire-Cracked Rock, and Nut Oil: Exploring the Origins of Pottery Making
on Grand Island. The Wisconsin Archeologist 90:132-149.
Skibo, J. M., T. Martin, E. Drake, and J. Franzen
2004 Gete Odena: Grand Island’s Post-Contact Occupation at Williams Landing.
Midcontinental Journal of Archaeology 29(2)167-190.
351

Smith, B. A.
1983 Faunal Identification for Seven Northern Michigan Archaeological Sites: EkdahlGoodreau site, Portage site, Halberg site, Wood site, Wood Extension site, Mt. McSauba
site, and Neff’s A.H. Map site. On file at the Office of the State Archaeologist, Lansing,
Michigan.
1996 Systems of Subsistence and Networks of Exchange in the Terminal Woodland and Early
Historic Periods in the Upper Great Lakes. Unpublished Doctoral thesis, Department of
Anthropology, Michigan State University, East Lansing.
2004 The Gill Net’s “Native Country:” The Inland Shore Fishery in the Northern Great Lake
Michigan Basin. In An Upper Great Lakes Archaeological Odyssey: Essays in Honor of
Charles E. Cleland, edited by W. A. Lovis, Cranbrook Institute of Science and Wayne
State University Press, Detroit.
Smith, B. D.
2001 Low-Level Food Production. Journal of Archaeological Research 9(1):1-43.
2007 The Ultimate Ecosystem Engineers. Science 315:1797-1798.
Smith, C. S.
2003 Hunter-Gatherer mobility, storage, and houses in a marginal environment: an example
from the mid-Holocene of Wyoming. Journal of Anthropological Archaeology 22:162189.
Snow, D.
1976 Archaeology of North America. Viking Press, New York.
Sobel, E. and G. Bettles
2000 Winter Hunger, Winter Myths: Subsistence and Risk and Mythology among the Klamath
and Modoc. Journal of Anthropological Archaeology 19:276-316.
Sokal, R. and F. Rohlf
1995 Biometery: The Principles and Practice of Statistics in Biological Research. W. H.
Freeman Company, New York
Speth, J. D.
1991 Nutritional Constraints and Late Glacial Adaptive Transformations: The Importance of
Non-Protein Energy Sources. In The Late Glacial in North-West Europe: Human
Adaptation and Environmental Change at the End of the Pleistocene, edited by N.
Barton, A. Roberts, and D. Roe, pp. 169-178. Council for British Archaeology Research
Report 77, London.

352

Terrell, J. E., J. P. Hart, S. Barut, N. Cellinese, A. Curet, T. Denham, C. M. Kusimba, K. Latinis,
R. Oka, J. Palka, M. E. D. Pohl, K. O. Pope, P. R. Williams, H. Haines, and J. E. Staller.
2003 Domesticated Landscapes: The Subsistence Ecology of Plant and Animal Domestication.
Journal of Archaeological Method and Theory 10:323-368.
Thompson, V.
2010 The Rhythm of Space-Time and the Making of Monuments and Places during the
Archaic. In Trend, Tradition, and Turmoil: What Happened to the Southeastern
Archaic, edited by D. Thomas and M. Sanger, pp. 217-228. American Museum of
Natural History, New York.
Thompson, V. and J. Turck
2009 Adaptive Cycles of Coastal Hunter-Gatherers. American Antiquity 74(2):255-278
Trigger, B. G.
1968 The Determinants of Settlement Patterns. In Settlement Archaeology, edited by K. C.
Chang, pp. 53-78. National Press, Palo Alto.
Trusler, S. and L. M. Johnson
2008 “Berry Patch” As a Kind of Place – The Ethnoecology of Black Huckleberry in
Northwestern Canada. Human Ecology 36:553-568.
University of Michigan Museum of Anthropology
nd
20DE1 Archaeological Site Files. On file at the University of Michigan Museum of
Anthropology, Ann Arbor.
nd

20ST1 Archaeological Site Files. On file at the University of Michigan Museum of
Anthropology, Ann Arbor.

Van Deelen, T., K. Pregitzer, J. Haufler
1996 A Comparison of Presettlement and Present-Day Forests in Two Northern Michigan Deer
Yards. American Midlands Naturalist 135(2):181-194.
Vennum, T.
1988 Wild Rice and the Ojibwa People. Minnesota Historical Society Press, St. Paul
Voss, E. G. and A. Reznicek
2012 Michigan Flora: A Guide to the Identification and Occurrence of the Native and
Naturalized Seed-Plants of the State (2 volumes). Cranbrook Institute of Science and the
University of Michigan Herbarium, Bloomfield Hills, Michigan.
Walker, B., L. Gunderson, A. Kinzig, C. Folke, S. Carpenter, L. Schultz
2006 A Handful of Heuristics and Some Propositions for Understanding Resilience in SocialEcological Systems. Ecology and Society 11(1):13. Online resource (last accessed
1/3/2011): http://www.ecologyandsociety.org/vol11/iss1/art13/

353

Warren, R. E. and D. L. Asch
2000 A Predictive Model of Archaeological Site Location in the Eastern Prairie Peninsula. In
Practical Applications of GIS for Archaeologists: A Predictive Modeling Toolkit, edited
by K. Wescott and R. J. Brandon, pgs 5-32. Taylor and Francis, Inc., Philadelphia
Weir, D. J.
1981 A Cultural Reource Survey of the Hiawatha National Forest. Commonwealth Associates,
Inc., Jackson, Michigan.
Wescott, K. L.
2000 Introduction. In Practical Applications of GIS for Archaeologists: A Predictive Modeling
Tool Kit, edited by K. L. Wescott and J. J. Brandon, pp. 1-4. Taylor and Francis, London.
Whallon, R.
2006 Social networks and information: Non-“utilitarian” mobility among hunter-gatherers.
Journal of Anthropological Archaeology 25:259-270.
Whitcomb, L. D. Parker, B. Carr, P. Gobster, and H. Schroeder
2002 Human preference for ecological units: patterns of dispersed campsites within landtype
associations on the Chippewa National Forest. In Proceedings of the 2001 Northeastern
Recreation Research Symposium, edited by S. Todd, pp. 429-434. Gen. Tech. Rep. NE289, USDA Forest Service, Northeastern Research Station, Newtown Square,
Pennsylvania.
Whitney, G.
1986 Relation of Michigan’s Presettlement Pine Forest to Substrate and Disturbance History.
Ecology 67(6):1548-1559.
Willey, G. R.
1953 Prehistoric Settlement Patterns in the Virú Valley, Peru. Bulletin 155. Bureau of
American Ethnology, Washington, D.C.
Wills, W. H.
1992 Plant Cultivation and the Evolution of Risk-Prone Economies in the Prehistoric American
Southwest. In Transitions to Agriculture in Prehistory, edited by A. Gebauer and T. D.
Price, pp. 153-176. Prehistory Press, Madison, Wisconsin
Winkler, M., A. Swain, J. Kutzbach
1986 Middle Holocene Dry Period in the Northern Midwestern United States: Lake Levels
and Pollen Stratigraphy. Quaternary Research 25:235-250.
Wong, D. and J. Lee
2005 Statistical Analysis of Geographic Information with ArcView GIS and ArcGIS. Wiley,
Inc., Hoboken, New Jersey

354

Woods, K. and M. Davis
1989 Paleoecology of Range Limits: Beech in the Upper Peninsula of Michigan. Ecology
70(3):681-696.
Yarnell, R. A.
1964 Aboriginal Relationships between Culture and Plant Life in the Upper Great Lakes.
Anthropological Papers, Museum of Anthropology, University of Michigan 23, Ann
Arbor
Zhang, Q., K. Pregitzer, D. Reed
1999 Catastrophic Disturbance in the Presettlement Forests of the Upper Peninsula of
Michigan. Canadian Journal of Forest Research 29(1):106-114.
2000 Historical Changes in the Forests of the Luce District of the Upper Peninsula of
Michigan. American Midlands Naturalist 143(1):94-110

355